{
  "total": 764,
  "topics": [
    {
      "id": 157925,
      "title": "DFDC Final Results",
      "comment_count": 39,
      "views": 0,
      "votes": -48
    },
    {
      "id": 121179,
      "title": "Welcome!",
      "comment_count": 28,
      "views": 0,
      "votes": 19
    },
    {
      "id": 121202,
      "title": "New to Machine Learning or Kaggle?",
      "comment_count": 20,
      "views": 0,
      "votes": 17
    },
    {
      "id": 121203,
      "title": "External Data Disclosure Thread",
      "comment_count": 416,
      "views": 0,
      "votes": 37
    },
    {
      "id": 121201,
      "title": "Looking for a Team Thread",
      "comment_count": 199,
      "views": 0,
      "votes": 17
    },
    {
      "id": 691173,
      "title": "Dataset Access",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 596535,
      "title": "Is there a chance to download the test set from Kaggle?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 554325,
      "title": "Invalid aws account ID",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 365842,
      "title": "what is the best way to deal with this size of data ?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 145840,
      "title": "79st Place solution: A Hotchpotch of EfficientNets",
      "comment_count": 14,
      "views": 0,
      "votes": 20
    },
    {
      "id": 435424,
      "title": "Enhancing the Accuracy of Deep Fake Detection Model",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 158158,
      "title": "3rd place solution and code",
      "comment_count": 5,
      "views": 0,
      "votes": 50
    },
    {
      "id": 121173,
      "title": "1 Million Fake Faces on Kaggle",
      "comment_count": 89,
      "views": 0,
      "votes": 216
    },
    {
      "id": 456588,
      "title": "Dataset Download",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 123665,
      "title": "70,000 Real Faces",
      "comment_count": 7,
      "views": 0,
      "votes": 19
    },
    {
      "id": 121455,
      "title": "Face Reconstruction from Voice using Generative Adversarial Networks",
      "comment_count": 6,
      "views": 0,
      "votes": 23
    },
    {
      "id": 390238,
      "title": "Linking the Dataset",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 121223,
      "title": "State of the Art. Deepfake Detection references.",
      "comment_count": 19,
      "views": 0,
      "votes": 144
    },
    {
      "id": 121332,
      "title": "Discrepancy in Winners licensing terms",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121181,
      "title": "Check this out",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121313,
      "title": "GAN and DeepFake resources",
      "comment_count": 29,
      "views": 0,
      "votes": 95
    },
    {
      "id": 352022,
      "title": "Screenshot permission?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 145721,
      "title": "1st place solution",
      "comment_count": 102,
      "views": 0,
      "votes": 339
    },
    {
      "id": 249421,
      "title": "DFDC preview dataset",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121716,
      "title": "FaceForensics++ experiment",
      "comment_count": 33,
      "views": 0,
      "votes": 39
    },
    {
      "id": 122786,
      "title": "70k real faces",
      "comment_count": 24,
      "views": 0,
      "votes": 53
    },
    {
      "id": 286606,
      "title": "AttributeError: module 'cv2' has no attribute 'videoCapture'",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 157983,
      "title": "1st Place Removed Solution - All Faces Are Real Team",
      "comment_count": 393,
      "views": 0,
      "votes": 393
    },
    {
      "id": 145749,
      "title": "Is deepfake detection a lost cause?",
      "comment_count": 13,
      "views": 0,
      "votes": 25
    },
    {
      "id": 132647,
      "title": "Compressed Video Action Recognition",
      "comment_count": 2,
      "views": 0,
      "votes": 22
    },
    {
      "id": 122328,
      "title": "Video loading optimization tips",
      "comment_count": 31,
      "views": 0,
      "votes": 115
    },
    {
      "id": 123176,
      "title": "Andrew Trask comment on DeepFakes detection",
      "comment_count": 10,
      "views": 0,
      "votes": 9
    },
    {
      "id": 121481,
      "title": "Extracting Audio with FFMPEG without Internet in Kernel",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 241980,
      "title": "Which videos are/are not deepfakes?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 191072,
      "title": "Entire dataset now available--how do I download it?",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121422,
      "title": "ffmpeg-python package ",
      "comment_count": 4,
      "views": 0,
      "votes": 7
    },
    {
      "id": 140625,
      "title": "Journey to 0.3. Public 64th Place Solution",
      "comment_count": 11,
      "views": 0,
      "votes": 20
    },
    {
      "id": 140364,
      "title": "5th (from 20th) place approach",
      "comment_count": 27,
      "views": 0,
      "votes": 100
    },
    {
      "id": 164470,
      "title": "Any one in need of Processed Dataset",
      "comment_count": 1,
      "views": 0,
      "votes": -10
    },
    {
      "id": 145643,
      "title": "8th in the previous public leaderboard - No private scores and Public scores are removed",
      "comment_count": 19,
      "views": 0,
      "votes": 12
    },
    {
      "id": 133348,
      "title": "CNN-generated images are surprisingly easy to spot...for now",
      "comment_count": 4,
      "views": 0,
      "votes": 7
    },
    {
      "id": 140236,
      "title": "Public 2nd (Private 30th) Place Solution",
      "comment_count": 220,
      "views": 0,
      "votes": 328
    },
    {
      "id": 121251,
      "title": "Deepfake Detection vs Face Anti-spoofing",
      "comment_count": 2,
      "views": 0,
      "votes": 16
    },
    {
      "id": 182826,
      "title": "New Deepfake Detection Competition",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121594,
      "title": "Top 15 Free Image Datasets for Facial Recognition",
      "comment_count": 30,
      "views": 0,
      "votes": 62
    },
    {
      "id": 142757,
      "title": "How to enhance deepfake so that it can change face in real time?",
      "comment_count": 4,
      "views": 0,
      "votes": -4
    },
    {
      "id": 121664,
      "title": "Test Videos present in Train  dataset with label ?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 145626,
      "title": "67th Place Solution: Kaggle Kernals Only + Code",
      "comment_count": 7,
      "views": 0,
      "votes": 18
    },
    {
      "id": 129521,
      "title": "Using AWS Sagemaker with Existing Kaggle Datasets",
      "comment_count": 10,
      "views": 0,
      "votes": 15
    },
    {
      "id": 171865,
      "title": "How to use a prebuilt Keras model with no internet connection?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 169025,
      "title": "DeepTrace API ",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 127749,
      "title": "Submission and kaggle gpu quota",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 158506,
      "title": "2nd Place Removed Solution -QH Team ",
      "comment_count": 17,
      "views": 0,
      "votes": 70
    },
    {
      "id": 167413,
      "title": "Is detecting a deepfake taxing on GPU?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 165375,
      "title": "How to improve face recognition under occlusion condition？",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 130963,
      "title": "Image features extraction tools & resources",
      "comment_count": 5,
      "views": 0,
      "votes": 16
    },
    {
      "id": 121194,
      "title": "wget train files",
      "comment_count": 38,
      "views": 0,
      "votes": 38
    },
    {
      "id": 126108,
      "title": "where are the labels for the public validation set?",
      "comment_count": 6,
      "views": 0,
      "votes": 0
    },
    {
      "id": 128726,
      "title": "Kornia Data Augmentation",
      "comment_count": 5,
      "views": 0,
      "votes": 23
    },
    {
      "id": 158244,
      "title": "Question about the logic used to decide what is valid or not",
      "comment_count": 12,
      "views": 0,
      "votes": 28
    },
    {
      "id": 158553,
      "title": "The people on reddit acting like they could do better",
      "comment_count": 7,
      "views": 0,
      "votes": 10
    },
    {
      "id": 160319,
      "title": "Other Datasets for Deepfakes",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 158790,
      "title": "Did any of prize-winning teams, or top teams, use external data?",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 158272,
      "title": "Open message to the Kaggle team",
      "comment_count": 4,
      "views": 0,
      "votes": 19
    },
    {
      "id": 158145,
      "title": "Lecture I gave on deepfake detection",
      "comment_count": 5,
      "views": 0,
      "votes": 13
    },
    {
      "id": 148907,
      "title": "A round of applause for Giba",
      "comment_count": 11,
      "views": 0,
      "votes": 28
    },
    {
      "id": 158770,
      "title": "Workshop on Media Forensics - CVPR2020 - happening today(15 June)",
      "comment_count": 0,
      "views": 0,
      "votes": 10
    },
    {
      "id": 128946,
      "title": "ffmpeg vs cv2 for frames extraction",
      "comment_count": 7,
      "views": 0,
      "votes": 0
    },
    {
      "id": 158167,
      "title": "92th place solution",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 158413,
      "title": "Reinstatement of 20th team in private LB",
      "comment_count": 0,
      "views": 0,
      "votes": 7
    },
    {
      "id": 156124,
      "title": "2 month passed...",
      "comment_count": 7,
      "views": 0,
      "votes": 30
    },
    {
      "id": 152987,
      "title": "When will the leaderboard be finalized?",
      "comment_count": 5,
      "views": 0,
      "votes": 21
    },
    {
      "id": 121184,
      "title": "What is a deepfake, explained",
      "comment_count": 8,
      "views": 0,
      "votes": 8
    },
    {
      "id": 156234,
      "title": "accessing data on gcs from tpu machines for tf2",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 145623,
      "title": "The Private LB has been revealed!",
      "comment_count": 84,
      "views": 0,
      "votes": 32
    },
    {
      "id": 145740,
      "title": "Top private LB solutions",
      "comment_count": 15,
      "views": 0,
      "votes": 16
    },
    {
      "id": 145965,
      "title": "27th place solution",
      "comment_count": 8,
      "views": 0,
      "votes": 18
    },
    {
      "id": 139255,
      "title": "repeating error during commit ??",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 129817,
      "title": "TPU in Kaggle kernel?",
      "comment_count": 69,
      "views": 0,
      "votes": 11
    },
    {
      "id": 145822,
      "title": "12th place in a nutshell",
      "comment_count": 8,
      "views": 0,
      "votes": 19
    },
    {
      "id": 121401,
      "title": "Fake Faces - Some Observations",
      "comment_count": 7,
      "views": 0,
      "votes": 16
    },
    {
      "id": 140443,
      "title": "Top public solutions collection",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 150307,
      "title": "Public test set labels?",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 140417,
      "title": "Next steps in the DFDC competition",
      "comment_count": 29,
      "views": 0,
      "votes": 27
    },
    {
      "id": 121679,
      "title": "[SOLVED] Can't load private Keras models?",
      "comment_count": 9,
      "views": 0,
      "votes": 5
    },
    {
      "id": 147901,
      "title": "Public leaderboard update",
      "comment_count": 7,
      "views": 0,
      "votes": 3
    },
    {
      "id": 122168,
      "title": "Why don't we make Deepfake model as private?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 148502,
      "title": "EOFError (len(self.chunkname) < 4) when tried to access training data file",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 145982,
      "title": "113th place solution. Kaggle Team, whats up!?",
      "comment_count": 7,
      "views": 0,
      "votes": 10
    },
    {
      "id": 146864,
      "title": "Our solution",
      "comment_count": 2,
      "views": 0,
      "votes": 19
    },
    {
      "id": 146024,
      "title": "Did anyone get medal with single-model?",
      "comment_count": 13,
      "views": 0,
      "votes": 6
    },
    {
      "id": 145639,
      "title": "What a nice surprise! ",
      "comment_count": 23,
      "views": 0,
      "votes": 63
    },
    {
      "id": 147151,
      "title": "Submitting after competition end.",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 145713,
      "title": "AMA Interview with Yifan Xie | Competition Winning Team | Chai Time Data Science Show",
      "comment_count": 8,
      "views": 0,
      "votes": 10
    },
    {
      "id": 147265,
      "title": "Manually Labelled Actor Database",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 145658,
      "title": "COLLECTION of TEAMS invalidated in stage 2",
      "comment_count": 11,
      "views": 0,
      "votes": 19
    },
    {
      "id": 146587,
      "title": "5th place got Disqualified!?",
      "comment_count": 5,
      "views": 0,
      "votes": 6
    },
    {
      "id": 145972,
      "title": "How to access data in Colab",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 145732,
      "title": "39th Private LB: small and simple",
      "comment_count": 7,
      "views": 0,
      "votes": 28
    },
    {
      "id": 145648,
      "title": "Share your Public/Private scores! ",
      "comment_count": 18,
      "views": 0,
      "votes": 4
    },
    {
      "id": 145720,
      "title": "Silver zone solo, no credits used.",
      "comment_count": 6,
      "views": 0,
      "votes": 18
    },
    {
      "id": 140829,
      "title": "EDIT) Private 1089th Place Approach (BIG DROP from public 7th)",
      "comment_count": 30,
      "views": 0,
      "votes": 71
    },
    {
      "id": 145980,
      "title": "New Competition Grand Master: Selim Seferbekov",
      "comment_count": 0,
      "views": 0,
      "votes": 10
    },
    {
      "id": 145719,
      "title": "Reasons some kernels might have failed",
      "comment_count": 12,
      "views": 0,
      "votes": 6
    },
    {
      "id": 147205,
      "title": "Deepfake Competition",
      "comment_count": 1,
      "views": 0,
      "votes": -6
    },
    {
      "id": 145681,
      "title": "So close yet so far",
      "comment_count": 9,
      "views": 0,
      "votes": 6
    },
    {
      "id": 145977,
      "title": "Audio features",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 145841,
      "title": "43rd place private LB solution",
      "comment_count": 5,
      "views": 0,
      "votes": 11
    },
    {
      "id": 145886,
      "title": "Thoughts about this competition",
      "comment_count": 7,
      "views": 0,
      "votes": 6
    },
    {
      "id": 145629,
      "title": "Congrats to All Faces Are Real ",
      "comment_count": 7,
      "views": 0,
      "votes": 10
    },
    {
      "id": 145655,
      "title": "Recommends for re-runing",
      "comment_count": 6,
      "views": 0,
      "votes": 13
    },
    {
      "id": 144972,
      "title": "Last day...",
      "comment_count": 3,
      "views": 0,
      "votes": 16
    },
    {
      "id": 146065,
      "title": "Can any body tell, why this happen to my submission?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 145645,
      "title": "All submissions failed",
      "comment_count": 1,
      "views": 0,
      "votes": 8
    },
    {
      "id": 123138,
      "title": "import face_recognition -- ModuleNotFoundError: No module named",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 145947,
      "title": "Public set size 4k Private Set Size 10k == Notebook timeout",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 145632,
      "title": "Congratulations to the winner, and to everyone",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 145506,
      "title": "Thanks to Kaggle, Amazon & Google",
      "comment_count": 1,
      "views": 0,
      "votes": 5
    },
    {
      "id": 145663,
      "title": "Public vs Private Leaderboard Scatter Plot",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 145636,
      "title": "Invalidated for stage 2 error",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 145768,
      "title": "why log loss, but no acc or other evaluation criteria",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 145736,
      "title": "Congrats to winners",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 145214,
      "title": "Deepfakes evolution",
      "comment_count": 0,
      "views": 0,
      "votes": 9
    },
    {
      "id": 140303,
      "title": "Public 34th/Private 47th place insights - Part1 (Inference)",
      "comment_count": 15,
      "views": 0,
      "votes": 41
    },
    {
      "id": 125472,
      "title": "Anyone with a slow computer have a competitive rank?",
      "comment_count": 14,
      "views": 0,
      "votes": 3
    },
    {
      "id": 140878,
      "title": "Shake up or not shake up ?",
      "comment_count": 20,
      "views": 0,
      "votes": 23
    },
    {
      "id": 140299,
      "title": "Mistakes",
      "comment_count": 25,
      "views": 0,
      "votes": 51
    },
    {
      "id": 140071,
      "title": "What a Journey!",
      "comment_count": 29,
      "views": 0,
      "votes": 48
    },
    {
      "id": 144523,
      "title": "Deepfake Detection models for detect COVID-19 (only question) ?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 142606,
      "title": "FaceForensics++ dataset",
      "comment_count": 1,
      "views": 0,
      "votes": 16
    },
    {
      "id": 138901,
      "title": "Your notebook tried to allocate more memory than available.",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 143952,
      "title": "Share Faces Dataset or images",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121528,
      "title": "Dlib library without internet",
      "comment_count": 10,
      "views": 0,
      "votes": 5
    },
    {
      "id": 142946,
      "title": "Is Separable Conv effective in DeepFake Detection? And why?",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 134426,
      "title": "A simple script to monitor your submission completion",
      "comment_count": 3,
      "views": 0,
      "votes": 11
    },
    {
      "id": 142223,
      "title": "Evaluation on synthetic videos",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 140191,
      "title": "Model Thread - What model/s did you use?",
      "comment_count": 66,
      "views": 0,
      "votes": 22
    },
    {
      "id": 141417,
      "title": "Anybody having issues with training efficient-net?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 141482,
      "title": "Submission disabled ",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 140975,
      "title": "why I can not see my rank now?",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 140835,
      "title": "Why RNN (like LSTM) didn't worked?",
      "comment_count": 14,
      "views": 0,
      "votes": 4
    },
    {
      "id": 141072,
      "title": "deepfake on realtime webcam",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 140467,
      "title": "Splitting data into folds based on facially similar groupings",
      "comment_count": 7,
      "views": 0,
      "votes": 4
    },
    {
      "id": 141056,
      "title": "Thanks",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 140495,
      "title": "Extracted Face Dataset",
      "comment_count": 4,
      "views": 0,
      "votes": 0
    },
    {
      "id": 138953,
      "title": "Tips. How to get top 10%(for now)",
      "comment_count": 20,
      "views": 0,
      "votes": 36
    },
    {
      "id": 140099,
      "title": "Where's the endgame?",
      "comment_count": 4,
      "views": 0,
      "votes": 8
    },
    {
      "id": 138979,
      "title": "Go Vladislav!",
      "comment_count": 15,
      "views": 0,
      "votes": 29
    },
    {
      "id": 140234,
      "title": "Plz allow submissions, so that we can just see our scores",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 140831,
      "title": "List of videos with voice alteration?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 140830,
      "title": "good bye DFDC",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 139805,
      "title": "LB under 0.2",
      "comment_count": 8,
      "views": 0,
      "votes": 8
    },
    {
      "id": 140146,
      "title": "Submission impossible?",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 140018,
      "title": "2 submissions left!",
      "comment_count": 3,
      "views": 0,
      "votes": 6
    },
    {
      "id": 140281,
      "title": "Request to ground truth answer of test videos",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 140309,
      "title": "Share your notebooks and models!",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 140228,
      "title": "Can we do late submissions ( Not counted towards prizes) ?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 140246,
      "title": "Time to unveil esoteric methods :)",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 138639,
      "title": "REMINDER: Select your Final 2 Submissions by 3/31",
      "comment_count": 16,
      "views": 0,
      "votes": 14
    },
    {
      "id": 140357,
      "title": "Model run-time vs public score",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 139924,
      "title": "Does private dataset contain pairs of (true, fake) from the same video??",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 140476,
      "title": "Hardware - Which hardware did you use?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 140450,
      "title": "What are people’s thoughts on fake/real split bias?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 140202,
      "title": "My latest two submissions have been stuck in pending for 3 hours",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 139452,
      "title": "Electricity Burning Challenge",
      "comment_count": 5,
      "views": 0,
      "votes": 11
    },
    {
      "id": 139518,
      "title": "Getting a submission scoring error",
      "comment_count": 23,
      "views": 0,
      "votes": 0
    },
    {
      "id": 140037,
      "title": "private leaderboard == public leaderboard?",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 140358,
      "title": "What will be the efficient generalization strategy?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 140043,
      "title": "Good Luck & Thanks ",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 140280,
      "title": "Request to ground truth answer of test videos",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 139986,
      "title": "If I press the submit button before 11:59pm UTC, what happens to the submissions allowed perday?",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 135631,
      "title": "Request for extension due to the COVID-19 ",
      "comment_count": 58,
      "views": 0,
      "votes": 117
    },
    {
      "id": 140223,
      "title": "22 days to go means?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 139892,
      "title": "I got 11 submission scoring errors in a row...",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 139377,
      "title": "No-holdout validation and potential surprises",
      "comment_count": 6,
      "views": 0,
      "votes": 10
    },
    {
      "id": 138249,
      "title": "Some last minute insights",
      "comment_count": 5,
      "views": 0,
      "votes": 13
    },
    {
      "id": 140096,
      "title": "Why predictions of my model so skewed towards fakes?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 139968,
      "title": "Time of submission",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 139917,
      "title": "No more submissions?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 139876,
      "title": "Time change problem ? ",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 139920,
      "title": "Submission Options for >16mb RAM",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 139367,
      "title": "Submission notebook etiquette",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 139756,
      "title": "Submission errors?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 140124,
      "title": "don't know if it is too late to ask",
      "comment_count": 1,
      "views": 0,
      "votes": -4
    },
    {
      "id": 139502,
      "title": "single image models - Average score worst?",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 139747,
      "title": "Always a Notebook Timeout error",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 139325,
      "title": "Error in Submission",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 139741,
      "title": "our notebook tried to allocate more memory than is available. It has restarted.",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 139699,
      "title": "Is everyone okay with external data capacity?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 139696,
      "title": "Will submissions with non commerical use (e.g. pretrained weights) still get a ranking?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 139667,
      "title": "No other way to add our model to the kaggle kernel?",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 139528,
      "title": "Submission Scoring Error",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 139604,
      "title": "Test set release after competition ends?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 139547,
      "title": "Good luck detecting this deep fake method",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 139039,
      "title": "Keras preprocess_input()",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 138887,
      "title": "Worries about the submissions running properly",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 137109,
      "title": "[Solved] is training loss goes to ~ 0 (1e-12) quickly ?",
      "comment_count": 12,
      "views": 0,
      "votes": 5
    },
    {
      "id": 138837,
      "title": "Video Reader with GPU",
      "comment_count": 1,
      "views": 0,
      "votes": 8
    },
    {
      "id": 135648,
      "title": "Leaderboard groupings",
      "comment_count": 31,
      "views": 0,
      "votes": 21
    },
    {
      "id": 139007,
      "title": "Dead line extension request",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 139064,
      "title": "Size of external data",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 138452,
      "title": "Trust CV? LB?",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 138728,
      "title": "What informations are we allowed to use about test data?",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 139174,
      "title": "Problem for extension of deadline?",
      "comment_count": 5,
      "views": 0,
      "votes": -9
    },
    {
      "id": 128479,
      "title": "late submissions",
      "comment_count": 7,
      "views": 0,
      "votes": null
    },
    {
      "id": 139068,
      "title": "rsconnect timeout",
      "comment_count": 0,
      "views": 0,
      "votes": -1
    },
    {
      "id": 138578,
      "title": "Any external dataset really allowed?",
      "comment_count": 0,
      "views": 0,
      "votes": 9
    },
    {
      "id": 138808,
      "title": "Time limit on the PRIVATE test set",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 137135,
      "title": "Is this competition non commercial?",
      "comment_count": 3,
      "views": 0,
      "votes": 3
    },
    {
      "id": 138318,
      "title": "What makes overfitting?",
      "comment_count": 1,
      "views": 0,
      "votes": 8
    },
    {
      "id": 138862,
      "title": "How do you plan on testing whether the resolution of the test video is 1/4 in size?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 138795,
      "title": "Someone help clarify these two rules for me? (external code/data)",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 138772,
      "title": "What the best strategy to fine tune pretrained models?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 138439,
      "title": "Unable to install FFMPEG without internet, please Help",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 138640,
      "title": "Reminder - no Extension!",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 138187,
      "title": "small and clean train set or big but noisy?",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 137467,
      "title": "Optimum Clipping",
      "comment_count": 7,
      "views": 0,
      "votes": 3
    },
    {
      "id": 134366,
      "title": "LRCN Implementation (PyTorch)",
      "comment_count": 27,
      "views": 0,
      "votes": 33
    },
    {
      "id": 138352,
      "title": "Commit failed with code 137",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 138179,
      "title": "Why submission csv not found????",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 138083,
      "title": "I can not do submission",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 138069,
      "title": "Why do failed submissions count against daily submissions?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 138116,
      "title": "GPU quota has been extended?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 138086,
      "title": "My model seems to stop running a public test...",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 136099,
      "title": "What is the reason of choosing Log loss for scoring?",
      "comment_count": 3,
      "views": 0,
      "votes": 4
    },
    {
      "id": 137955,
      "title": "Is Kaggle team considering Extension",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 137963,
      "title": "Success with sequence models??",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 137917,
      "title": "Is this a fake?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 137898,
      "title": "Check external data limit in working notebook?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 137756,
      "title": "May I ask you a question about my submission?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 137221,
      "title": "Kaggle Error when submitted",
      "comment_count": 3,
      "views": 0,
      "votes": 3
    },
    {
      "id": 131274,
      "title": "Best single model score",
      "comment_count": 139,
      "views": 0,
      "votes": 28
    },
    {
      "id": 136966,
      "title": "Deepfake detection image benchmark",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 129040,
      "title": "How did we get our current score(0.38504)",
      "comment_count": 39,
      "views": 0,
      "votes": 71
    },
    {
      "id": 136924,
      "title": "FAKE videos' frames",
      "comment_count": 14,
      "views": 0,
      "votes": 1
    },
    {
      "id": 133316,
      "title": "The rules seem vague. Kaggle needs to clarify them ASAP",
      "comment_count": 36,
      "views": 0,
      "votes": 26
    },
    {
      "id": 137290,
      "title": "What does your offline inference pipeline look like?",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 136869,
      "title": "Anyone's tried Bojan's dataset?",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 136029,
      "title": "Curious, what has been faked? Is actor different?",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 136985,
      "title": "Validation loss is not enough to choose a model ",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 137687,
      "title": "Breaking the best submission",
      "comment_count": 0,
      "views": 0,
      "votes": -8
    },
    {
      "id": 133696,
      "title": "Video reader",
      "comment_count": 11,
      "views": 0,
      "votes": 8
    },
    {
      "id": 137511,
      "title": "Interesting challenge! :D",
      "comment_count": 0,
      "views": 0,
      "votes": -5
    },
    {
      "id": 134420,
      "title": "Other useful datasets finished",
      "comment_count": 5,
      "views": 0,
      "votes": 17
    },
    {
      "id": 136942,
      "title": "Are we allowed to present our approach/methodology to a workshop before the competion ends?",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 137138,
      "title": "Trouble with training on Colab",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 137036,
      "title": "Have any tried Inception Score(or similar metric)?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 136923,
      "title": "COVID-19: Postponing the end of the competition",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 136117,
      "title": "TPU vs GPU ?",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 136839,
      "title": "what is this error? ",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 136934,
      "title": "FAKE videos' frames",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 136922,
      "title": "Notebook Exceeded Allowed Compute",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 136871,
      "title": "how to print image with color",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 134733,
      "title": "The fight for generalization",
      "comment_count": 11,
      "views": 0,
      "votes": 30
    },
    {
      "id": 134485,
      "title": "More timeout issues",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 131843,
      "title": "[Taking Request] Notebook Publish",
      "comment_count": 48,
      "views": 0,
      "votes": 25
    },
    {
      "id": 132785,
      "title": "LB is always 0.2 higher than the local validation score",
      "comment_count": 15,
      "views": 0,
      "votes": 1
    },
    {
      "id": 136802,
      "title": "Help with accessing audio data",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 135857,
      "title": "Submission CSV not found",
      "comment_count": 14,
      "views": 0,
      "votes": null
    },
    {
      "id": 136151,
      "title": "About the requirements",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 136146,
      "title": "Real faces datasets listed anywhere?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 135756,
      "title": "Should I release my data set?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 135799,
      "title": "Analysis of submission.csv",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 135441,
      "title": "Class Activations on ResNext50",
      "comment_count": 2,
      "views": 0,
      "votes": 12
    },
    {
      "id": 135881,
      "title": "Regarding Submission",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 134456,
      "title": "best frames_per_video for prediction",
      "comment_count": 16,
      "views": 0,
      "votes": 10
    },
    {
      "id": 134608,
      "title": "Detection surfaces",
      "comment_count": 3,
      "views": 0,
      "votes": 38
    },
    {
      "id": 135867,
      "title": "allowed runtime cap for the competition",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 135543,
      "title": "Slide Algorithm - might be a new era in deep learning",
      "comment_count": 3,
      "views": 0,
      "votes": 4
    },
    {
      "id": 135306,
      "title": "Submitting a published kernel.",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 135630,
      "title": "Strategy to select final 2 submissions?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 135319,
      "title": "Efficient net Converge issues",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 135593,
      "title": "Cuda error",
      "comment_count": 7,
      "views": 0,
      "votes": 0
    },
    {
      "id": 133460,
      "title": "A different approach",
      "comment_count": 1,
      "views": 0,
      "votes": 7
    },
    {
      "id": 135134,
      "title": "Analyzing Errors (the images)",
      "comment_count": 1,
      "views": 0,
      "votes": 18
    },
    {
      "id": 135453,
      "title": "The way to calculate the percentage of prediction?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 134895,
      "title": "Is this what overfitting looks like?",
      "comment_count": 16,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121183,
      "title": "Useful Papers",
      "comment_count": 25,
      "views": 0,
      "votes": 144
    },
    {
      "id": 135633,
      "title": "xception vs resnet result",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 135565,
      "title": "Download data from GCP ",
      "comment_count": 8,
      "views": 0,
      "votes": null
    },
    {
      "id": 134983,
      "title": "Why always submit error？",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 135470,
      "title": "An interesting reddit AMA relevant to deepfakes and their implications",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 128954,
      "title": "Other useful datasets",
      "comment_count": 80,
      "views": 0,
      "votes": 71
    },
    {
      "id": 135510,
      "title": "Submission Error ",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 134936,
      "title": "Analyzing Errors (the numbers)",
      "comment_count": 1,
      "views": 0,
      "votes": 14
    },
    {
      "id": 135448,
      "title": "Tuning Efficient Net and Resenet",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 134352,
      "title": "Funny FAKE videos",
      "comment_count": 6,
      "views": 0,
      "votes": 13
    },
    {
      "id": 123737,
      "title": "Will the private test execute in exactly the same environment?",
      "comment_count": 3,
      "views": 0,
      "votes": 3
    },
    {
      "id": 135327,
      "title": "Not Able to create new notebook in Deepfake Detection Challange",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 134845,
      "title": "The magic of 0.69314",
      "comment_count": 11,
      "views": 0,
      "votes": 1
    },
    {
      "id": 134970,
      "title": "ImageDataGenerators for Keras",
      "comment_count": 7,
      "views": 0,
      "votes": null
    },
    {
      "id": 134897,
      "title": "Can the organizers re-run the submissions with kaggle error?",
      "comment_count": 1,
      "views": 0,
      "votes": 8
    },
    {
      "id": 134216,
      "title": "Paperswithcode for deepfake",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 135384,
      "title": "Topic Title",
      "comment_count": 4,
      "views": 0,
      "votes": -6
    },
    {
      "id": 135105,
      "title": "Have sub errors resolved ",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 134990,
      "title": "Option for more detailed error messages for Kaggle scripts?",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 135147,
      "title": "What does 'bounded away' mean during Evaluation?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 134921,
      "title": "A fun challenge for beginners: Predict if an image is GAN generated",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 132648,
      "title": "Inching towards 2000+ teams",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 122254,
      "title": "AWS Promotional Credit Distribution Guidelines for DFDC",
      "comment_count": 137,
      "views": 0,
      "votes": 91
    },
    {
      "id": 134998,
      "title": "The Fake Price of Bel-Air",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 134774,
      "title": "The kaggle error makes me speechless",
      "comment_count": 11,
      "views": 0,
      "votes": 1
    },
    {
      "id": 134841,
      "title": "Access to FaceForensics++ dataset as a non-researcher?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 134934,
      "title": "How to achieve real-time deep face?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 132700,
      "title": "How to train with all dataset?",
      "comment_count": 24,
      "views": 0,
      "votes": 15
    },
    {
      "id": 133748,
      "title": "What about clipping?",
      "comment_count": 14,
      "views": 0,
      "votes": 1
    },
    {
      "id": 134308,
      "title": "Does cleaning the dataset damage test predictions?",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 134817,
      "title": "Using Blink Rate as Detection?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 134274,
      "title": "Ensemble Benefit !?!",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 134359,
      "title": "Let's see how fast this thing overfits",
      "comment_count": 1,
      "views": 0,
      "votes": 9
    },
    {
      "id": 134625,
      "title": "Submission error: Notebook Threw Exception",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 134167,
      "title": "Interesting observation",
      "comment_count": 3,
      "views": 0,
      "votes": 8
    },
    {
      "id": 134427,
      "title": "How to use multi-gpu?",
      "comment_count": 7,
      "views": 0,
      "votes": 1
    },
    {
      "id": 133975,
      "title": "All faces?",
      "comment_count": 13,
      "views": 0,
      "votes": 2
    },
    {
      "id": 132172,
      "title": "Teaming or not?",
      "comment_count": 9,
      "views": 0,
      "votes": 3
    },
    {
      "id": 133225,
      "title": "|| processing",
      "comment_count": 10,
      "views": 0,
      "votes": 4
    },
    {
      "id": 134428,
      "title": "Do I have to disclose my own model that I use?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 134295,
      "title": "<CNN-generated images are surprisingly easy to spot...for now>",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 134065,
      "title": "Submission error",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 131060,
      "title": "Simple Classifier Baseline (LB .489)",
      "comment_count": 30,
      "views": 0,
      "votes": 28
    },
    {
      "id": 132412,
      "title": "Transformer-like models?",
      "comment_count": 3,
      "views": 0,
      "votes": 6
    },
    {
      "id": 130903,
      "title": "How to do data augmentation for face recognition?",
      "comment_count": 12,
      "views": 0,
      "votes": 11
    },
    {
      "id": 133421,
      "title": "Reminder from the organizers",
      "comment_count": 10,
      "views": 0,
      "votes": 17
    },
    {
      "id": 134197,
      "title": "Using ffmpeg cannot \"commit to competition\" ?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 134208,
      "title": "GPU utilization",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 131157,
      "title": "A question about submissions",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 132890,
      "title": "how to offline install pyav?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 132251,
      "title": "NVIDIA DALI fails at submission",
      "comment_count": 12,
      "views": 0,
      "votes": 7
    },
    {
      "id": 134122,
      "title": "check accuracy on future epochs",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 133954,
      "title": "Basics question on using facenet_pytorch (in a loop) with efficient memory",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 127824,
      "title": "Submission Errors Solutions",
      "comment_count": 8,
      "views": 0,
      "votes": 26
    },
    {
      "id": 133534,
      "title": "Difference between training on jpg and png faces?",
      "comment_count": 11,
      "views": 0,
      "votes": 2
    },
    {
      "id": 133355,
      "title": "How to become Not Eligible for Prize?",
      "comment_count": 10,
      "views": 0,
      "votes": 0
    },
    {
      "id": 133875,
      "title": "CV2 and MTCNN takes ~10 hours on kaggle kernel?",
      "comment_count": 11,
      "views": 0,
      "votes": 0
    },
    {
      "id": 134280,
      "title": "Last Call on anybody needing some cloud credits ASAP",
      "comment_count": 0,
      "views": 0,
      "votes": -2
    },
    {
      "id": 134158,
      "title": "Should submission notebook include training script?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 133721,
      "title": "Has anyone got benefit from audio?",
      "comment_count": 3,
      "views": 0,
      "votes": 4
    },
    {
      "id": 134093,
      "title": "Classy Vision",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 133317,
      "title": "Is this a useful exercise?",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 133684,
      "title": "Face-detection based model limitations?",
      "comment_count": 8,
      "views": 0,
      "votes": 3
    },
    {
      "id": 132858,
      "title": "Questions about Private Test Set",
      "comment_count": 3,
      "views": 0,
      "votes": 22
    },
    {
      "id": 130759,
      "title": "A simple approach to classifying videos",
      "comment_count": 22,
      "views": 0,
      "votes": 17
    },
    {
      "id": 134077,
      "title": "Pre-processing Image Ideas",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 133654,
      "title": "Cropping Faces",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 132590,
      "title": "My notebook runs on Jupyter but throws up an error in Kaggle kernel",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 133999,
      "title": "Memory Usage",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 132649,
      "title": "Your Notebook cannot use non-standard datasets in this competition ",
      "comment_count": 9,
      "views": 0,
      "votes": 16
    },
    {
      "id": 133623,
      "title": "Public Validation Set vs Public Test Set 400 vs 4000",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 129871,
      "title": "Can we use TPU to train models with PyTorch?",
      "comment_count": 9,
      "views": 0,
      "votes": 8
    },
    {
      "id": 133717,
      "title": "Validation charts",
      "comment_count": 0,
      "views": 0,
      "votes": -1
    },
    {
      "id": 133197,
      "title": "bad validation and \"jumpy\" predictions ...why?",
      "comment_count": 14,
      "views": 0,
      "votes": 1
    },
    {
      "id": 133677,
      "title": "Entries on LB != Number of Submission ",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 133640,
      "title": "Two submissions failed, not sure if I'm out of competition",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 133354,
      "title": "Model not training",
      "comment_count": 8,
      "views": 0,
      "votes": 1
    },
    {
      "id": 126691,
      "title": "Why you should NOT use the provided data chunks as valid/train split",
      "comment_count": 24,
      "views": 0,
      "votes": 63
    },
    {
      "id": 133199,
      "title": "Is efficientnet allowed?",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 133318,
      "title": "The right way to ensemble models",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 133532,
      "title": "submission process",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 132328,
      "title": "[just for fun] Politicians please do/don't wear shirts with faces? ",
      "comment_count": 3,
      "views": 0,
      "votes": 5
    },
    {
      "id": 132101,
      "title": "Ensemble suggestions?",
      "comment_count": 26,
      "views": 0,
      "votes": 10
    },
    {
      "id": 132932,
      "title": "A better way to sort out training examples?",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 132885,
      "title": "Score difference between test and LB results",
      "comment_count": 9,
      "views": 0,
      "votes": 2
    },
    {
      "id": 130762,
      "title": "Can anyone share face detected bbox's?",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 132205,
      "title": "Reminder March 3 Entry, Team Merger & External Data Deadline",
      "comment_count": 14,
      "views": 0,
      "votes": 7
    },
    {
      "id": 133281,
      "title": "Detecting faces",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 132222,
      "title": "Leader Board Scoring Anomaly x3 Identical scores that shouldn't be??",
      "comment_count": 7,
      "views": 0,
      "votes": null
    },
    {
      "id": 131842,
      "title": "Has anyone done an analysis on how many real frames are found in a labelled 'FAKE' video?",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 133177,
      "title": "The mystery of Kaggle error and 10 hour kernel",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 132366,
      "title": "Different targets in the dialog videos",
      "comment_count": 7,
      "views": 0,
      "votes": 7
    },
    {
      "id": 132508,
      "title": "New number one",
      "comment_count": 8,
      "views": 0,
      "votes": 6
    },
    {
      "id": 129832,
      "title": "All actors",
      "comment_count": 21,
      "views": 0,
      "votes": 66
    },
    {
      "id": 133053,
      "title": "Failing to commit due to pip? ",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 132991,
      "title": "Competition Theme Song (DeepFake FaceSwap Music Video)",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 133015,
      "title": "[Solved] Input dataset size limit + Pytorch 1.4 - Update Notebook/Smaller whl?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 132538,
      "title": "custom dataset generator",
      "comment_count": 2,
      "views": 0,
      "votes": 0
    },
    {
      "id": 130800,
      "title": "Some insights",
      "comment_count": 12,
      "views": 0,
      "votes": 24
    },
    {
      "id": 124127,
      "title": "Validation strategies?",
      "comment_count": 27,
      "views": 0,
      "votes": 27
    },
    {
      "id": 130502,
      "title": "Fake Audio Detection Broad Guidance",
      "comment_count": 11,
      "views": 0,
      "votes": 11
    },
    {
      "id": 132799,
      "title": "your notebook cannot use internet access for this competition",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 132748,
      "title": "Jumping val loss",
      "comment_count": 3,
      "views": 0,
      "votes": 0
    },
    {
      "id": 132666,
      "title": "Frame-by-frame training ?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 132461,
      "title": "Anybody need some Cloud Credits for next week?",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 126623,
      "title": "Memory leaks in Kaggle's kernel",
      "comment_count": 18,
      "views": 0,
      "votes": 6
    },
    {
      "id": 131778,
      "title": "Videos with multiple actors",
      "comment_count": 5,
      "views": 0,
      "votes": 4
    },
    {
      "id": 124728,
      "title": "CV vs LB",
      "comment_count": 80,
      "views": 0,
      "votes": 34
    },
    {
      "id": 126156,
      "title": "Using AWS for DFDC",
      "comment_count": 15,
      "views": 0,
      "votes": 23
    },
    {
      "id": 131981,
      "title": "Training in Kaggle vs Colab vs SageMaker (ml.p2.xlarge)",
      "comment_count": 18,
      "views": 0,
      "votes": 3
    },
    {
      "id": 132625,
      "title": "Product Feedback for Kaggle: Display Run Time of Submission Notebook",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 132684,
      "title": "Faceextraction: Best models / strategies MTCNN/S3FD/BlazeFace",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 132082,
      "title": "Let's talk about GPU Spec & Training Time",
      "comment_count": 20,
      "views": 0,
      "votes": 4
    },
    {
      "id": 129866,
      "title": "Dealing with unbalanced data?",
      "comment_count": 16,
      "views": 0,
      "votes": 6
    },
    {
      "id": 132273,
      "title": "Submission error: Exceeded allowed compute",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 132437,
      "title": "Clustering-based Validation Scheme",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 132603,
      "title": "Google drive & cloud",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 131695,
      "title": "frame counts",
      "comment_count": 8,
      "views": 0,
      "votes": 5
    },
    {
      "id": 131978,
      "title": "How to upload the externally trained model?",
      "comment_count": 9,
      "views": 0,
      "votes": 1
    },
    {
      "id": 128864,
      "title": "Setup Google Cloud Instance for DFDC",
      "comment_count": 15,
      "views": 0,
      "votes": 25
    },
    {
      "id": 132196,
      "title": "\"Unable to find submission\" error while Notebook still running...",
      "comment_count": 7,
      "views": 0,
      "votes": 3
    },
    {
      "id": 125313,
      "title": "Request Free Cloud TPU Quota!",
      "comment_count": 109,
      "views": 0,
      "votes": 59
    },
    {
      "id": 130224,
      "title": "face_recognition not recognizes faces in certain videos",
      "comment_count": 12,
      "views": 0,
      "votes": 4
    },
    {
      "id": 131863,
      "title": "Manoj Tiwari has used artificial intelligence techniques to make it look like he said things he didn’t say",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 132154,
      "title": "Is it allowed to use YouTube videos for training?",
      "comment_count": 0,
      "views": 0,
      "votes": 8
    },
    {
      "id": 132466,
      "title": "Real videos only?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 129140,
      "title": "Tips to deal with overfitting?",
      "comment_count": 13,
      "views": 0,
      "votes": 2
    },
    {
      "id": 132235,
      "title": "Not displaying errors does not prevent probing.",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 132138,
      "title": "Submission is failing with the following error: Notebook Exceeded Allowed Compute",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 132266,
      "title": "Missing video in data dfdc_train_part_35",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 132005,
      "title": "Your notebook cannot use non-standard datasets in this competiton",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 124713,
      "title": "Private Test Set: Format and length of the videos ?",
      "comment_count": 6,
      "views": 0,
      "votes": 13
    },
    {
      "id": 132162,
      "title": "Human-'Robot/AI' Interactive RL for deepfake detection.",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 131794,
      "title": "Effective way to store/load video frames?",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 131986,
      "title": "Yet another pretrained model",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 131877,
      "title": "Loss is not decreasing ",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 131532,
      "title": "Submission CSV Not Found，but resubmit can get score",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 129874,
      "title": "Through the bubble",
      "comment_count": 18,
      "views": 0,
      "votes": 6
    },
    {
      "id": 131835,
      "title": "Submission CSV NOT Found please help me..",
      "comment_count": 20,
      "views": 0,
      "votes": null
    },
    {
      "id": 131688,
      "title": "Pls dont post high scoring kernel ,just strategy still fine",
      "comment_count": 3,
      "views": 0,
      "votes": 4
    },
    {
      "id": 131121,
      "title": "Is using imagenet pretrained models allowed?",
      "comment_count": 6,
      "views": 0,
      "votes": 10
    },
    {
      "id": 131566,
      "title": "model usage",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 130785,
      "title": "dlip package - speed",
      "comment_count": 8,
      "views": 0,
      "votes": 2
    },
    {
      "id": 131216,
      "title": "BAD Training Data!",
      "comment_count": 16,
      "views": 0,
      "votes": 1
    },
    {
      "id": 131790,
      "title": "Test videos",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 123707,
      "title": "Resized first frames and labels dataset",
      "comment_count": 19,
      "views": 0,
      "votes": 23
    },
    {
      "id": 129015,
      "title": "We need to read audio from the files.",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 131831,
      "title": "Do open source software need to be declared?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 131647,
      "title": "Time limit",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 126955,
      "title": "Errors while submitting Audio only kernel",
      "comment_count": 5,
      "views": 0,
      "votes": 7
    },
    {
      "id": 131431,
      "title": "Committed kernel - works, submit for competition - notebook timeout",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 131150,
      "title": "Is it necessary to align detected faces rather than only cropping them?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 131708,
      "title": "uploading output file on kaggle kernel for submission",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 131553,
      "title": "Submission error even though commit works",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 131698,
      "title": "best score using publicly available data",
      "comment_count": 0,
      "views": 0,
      "votes": -1
    },
    {
      "id": 131446,
      "title": "Add Extra Package?",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 131315,
      "title": "Today on Arxiv...",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 125612,
      "title": "What does \"Kaggle Error\" means?",
      "comment_count": 19,
      "views": 0,
      "votes": 6
    },
    {
      "id": 131273,
      "title": "resize and detect or the other way around?",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 131286,
      "title": "Downloading Data on AWS",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 130201,
      "title": "problem with submission",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 130635,
      "title": "face detection method for inference",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121861,
      "title": "~8% audio altered",
      "comment_count": 40,
      "views": 0,
      "votes": 52
    },
    {
      "id": 131313,
      "title": "Are test labels wrong?",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 131425,
      "title": "Merger Deadline",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 128702,
      "title": "[placeholder] pytorch starter kit",
      "comment_count": 36,
      "views": 0,
      "votes": 36
    },
    {
      "id": 131377,
      "title": "Submission CSV not found.",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 130124,
      "title": "Ratio of True to Fake in Test set",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 131280,
      "title": "Submission Falied",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 131301,
      "title": "how to create a custom python??",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 121695,
      "title": "First steps to get started",
      "comment_count": 39,
      "views": 0,
      "votes": 95
    },
    {
      "id": 131209,
      "title": "transfer learning vs model from scratch",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 129663,
      "title": "Resnet + LSTM",
      "comment_count": 20,
      "views": 0,
      "votes": 17
    },
    {
      "id": 131083,
      "title": "AWS Limit increase - Pending amazon action",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 131200,
      "title": "Adding data fails",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 131166,
      "title": "astwr",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 130581,
      "title": "Maximum Submission Constraint",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 131141,
      "title": "No R ?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 124583,
      "title": "Useful Model Structures",
      "comment_count": 20,
      "views": 0,
      "votes": 56
    },
    {
      "id": 130915,
      "title": "CV by actors?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 126940,
      "title": "paralellize pre-processing ...how? any tips?",
      "comment_count": 14,
      "views": 0,
      "votes": 5
    },
    {
      "id": 131035,
      "title": "How do I upload trained weights for my model for inference?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 130640,
      "title": "Image Normalization",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 130718,
      "title": "Do the submitted kernels count towards the weekly GPU quota?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 130754,
      "title": "Unable to pip install imageio-ffmpeg",
      "comment_count": 7,
      "views": 0,
      "votes": null
    },
    {
      "id": 129770,
      "title": "What kind of hardware specs are top teams using?",
      "comment_count": 10,
      "views": 0,
      "votes": 2
    },
    {
      "id": 130637,
      "title": "What infra are you using?",
      "comment_count": 8,
      "views": 0,
      "votes": 1
    },
    {
      "id": 130475,
      "title": "Data bucket on GCS",
      "comment_count": 10,
      "views": 0,
      "votes": 1
    },
    {
      "id": 130860,
      "title": "Audio Classification",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 130415,
      "title": "how to deal with audio and avoid submission error?",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 127492,
      "title": "Dealing with multiple faces detection",
      "comment_count": 8,
      "views": 0,
      "votes": 4
    },
    {
      "id": 130472,
      "title": "Submission Scoring error",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 128919,
      "title": "Major difference between CV and LB",
      "comment_count": 32,
      "views": 0,
      "votes": 17
    },
    {
      "id": 130666,
      "title": "Any success with Test Time Augmentation?",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 130680,
      "title": "Automated Machine Learning Tools meaning?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 130440,
      "title": "What do we need to predict?",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 130688,
      "title": "error No space left on device",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 130643,
      "title": "Some ideas to try",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 130579,
      "title": "How is the 1GB limit measured?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 123000,
      "title": "Possible Mislabelled data?",
      "comment_count": 7,
      "views": 0,
      "votes": 5
    },
    {
      "id": 130536,
      "title": "Low GPU utilization in kernel",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 127482,
      "title": "Suggestions for cropping face from video",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 130535,
      "title": "Notebook commit throwing error",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 130534,
      "title": "Output zip file not visible",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 130438,
      "title": "Beginner Queries",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 123457,
      "title": "Training Models on AWS Spot Instances",
      "comment_count": 6,
      "views": 0,
      "votes": 9
    },
    {
      "id": 130473,
      "title": "Submission Timed Out",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 130149,
      "title": "TPU for submission",
      "comment_count": 3,
      "views": 0,
      "votes": 3
    },
    {
      "id": 129949,
      "title": "CV keeps jumping around during training",
      "comment_count": 10,
      "views": 0,
      "votes": 2
    },
    {
      "id": 130444,
      "title": "Error on commit",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 130155,
      "title": "wondering what program you use to go over the images? Linux",
      "comment_count": 8,
      "views": 0,
      "votes": null
    },
    {
      "id": 129290,
      "title": "Hand Labeling Training Set...",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 129640,
      "title": "Real vs Fake Faces: Should I make a dataset or in-class competition?",
      "comment_count": 4,
      "views": 0,
      "votes": 7
    },
    {
      "id": 125375,
      "title": "People with Multiple IDs on the LB",
      "comment_count": 7,
      "views": 0,
      "votes": 24
    },
    {
      "id": 127695,
      "title": "Big shakeup at top",
      "comment_count": 38,
      "views": 0,
      "votes": 6
    },
    {
      "id": 128797,
      "title": "Is pseudo-labeling allowed?",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 130111,
      "title": "Pytorch MTCNN",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 126233,
      "title": "Not all frames from fake videos are fake !",
      "comment_count": 30,
      "views": 0,
      "votes": 13
    },
    {
      "id": 128150,
      "title": "Frustrating \"cloud\" training",
      "comment_count": 10,
      "views": 0,
      "votes": 4
    },
    {
      "id": 129957,
      "title": "How long submission takes ?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 129791,
      "title": "Which method is used for voice manipulation?",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 129537,
      "title": "Will NeuralTextures be on the test? :-)",
      "comment_count": 12,
      "views": 0,
      "votes": 2
    },
    {
      "id": 129935,
      "title": "What would be the best possible achievable score???",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 129675,
      "title": "We also have the definition of CHEAP FAKES",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 129860,
      "title": "We need to use the face_recognition to recognize the faces first.",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 122257,
      "title": "Help with overfitting",
      "comment_count": 13,
      "views": 0,
      "votes": 2
    },
    {
      "id": 129470,
      "title": "Identify fake videos by people's movements",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 129222,
      "title": "Submission issue: succesful run on public set, but immediate (failed?) run on private set",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 124417,
      "title": "Hash conclusions",
      "comment_count": 24,
      "views": 0,
      "votes": 69
    },
    {
      "id": 129710,
      "title": "Notebook Threw Exception",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 129276,
      "title": "How to use GPU with pytorch on my local machine",
      "comment_count": 8,
      "views": 0,
      "votes": 1
    },
    {
      "id": 127338,
      "title": "Public test face recognition videos fundamentally different from training/public val? Issues with face_recognition",
      "comment_count": 4,
      "views": 0,
      "votes": 6
    },
    {
      "id": 129354,
      "title": "Are  we  the  real  human  live  in the  real  world ?",
      "comment_count": 8,
      "views": 0,
      "votes": 3
    },
    {
      "id": 129108,
      "title": "Submission process understanding",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 129173,
      "title": "Can you please increase the number of submissions per day before the end?",
      "comment_count": 17,
      "views": 0,
      "votes": 3
    },
    {
      "id": 129444,
      "title": "Unreproducible result between kaggle and AWS Sagemaker",
      "comment_count": 7,
      "views": 0,
      "votes": 1
    },
    {
      "id": 129204,
      "title": "Can be useful for someone who doesn't have a powerful PC and/or isn't able to afford a cloud service!",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 128925,
      "title": "AWS credits  for DFDC",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 129367,
      "title": "use a pretrained model",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 124961,
      "title": " Facebook bans 'deepfake' videos in run-up to US election",
      "comment_count": 3,
      "views": 0,
      "votes": 6
    },
    {
      "id": 129350,
      "title": "load all data on kaggle kernel",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121441,
      "title": "DFDC:  Insights about the Dataset creation, baselines...",
      "comment_count": 4,
      "views": 0,
      "votes": 52
    },
    {
      "id": 126939,
      "title": "Deepfake Challenge Presentation : Status-Tips-Discussion Summary *Continuously Updating*",
      "comment_count": 6,
      "views": 0,
      "votes": 8
    },
    {
      "id": 129059,
      "title": "Few more useful papers on the subject of deepfakes",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 129179,
      "title": "working without using notebook",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 129095,
      "title": "Questions from a beginner",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 129109,
      "title": "Denoising frames",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 123080,
      "title": "Metadata update coming soon!",
      "comment_count": 10,
      "views": 0,
      "votes": 19
    },
    {
      "id": 128646,
      "title": "Which Instance type Should i use ?",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 127702,
      "title": "Submission finishes so fast...",
      "comment_count": 11,
      "views": 0,
      "votes": 7
    },
    {
      "id": 128650,
      "title": "Another data leak in public test set?",
      "comment_count": 5,
      "views": 0,
      "votes": 3
    },
    {
      "id": 129044,
      "title": "Idea from a noob",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 129000,
      "title": "Mismatch between local test and public test result",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 128533,
      "title": "commit failed without module named \"dlib\"!!!!",
      "comment_count": 7,
      "views": 0,
      "votes": 1
    },
    {
      "id": 128779,
      "title": "PERMISSION_DENIED error",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 128380,
      "title": "Submission failures investigation",
      "comment_count": 5,
      "views": 0,
      "votes": 11
    },
    {
      "id": 128291,
      "title": "All the faces of the train sample",
      "comment_count": 9,
      "views": 0,
      "votes": 7
    },
    {
      "id": 126729,
      "title": "Submission Takes Longer Than Commit?",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 121206,
      "title": "Isn't this going to make deepfakes much better?",
      "comment_count": 10,
      "views": 0,
      "votes": 27
    },
    {
      "id": 122917,
      "title": "Data leak in metadata",
      "comment_count": 9,
      "views": 0,
      "votes": 58
    },
    {
      "id": 128413,
      "title": "what is the Best LB score using only Face crops and no audio analysis",
      "comment_count": 6,
      "views": 0,
      "votes": 2
    },
    {
      "id": 128538,
      "title": "How much money will you get after taking out TAX?",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 128645,
      "title": "[pytorch] ohem loss implementation",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 128279,
      "title": "Submission csv not found",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 128562,
      "title": "What does this mean \" (unless otherwise agreed to by Competition Sponsor\"?",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121348,
      "title": "Fill  the same value simply",
      "comment_count": 14,
      "views": 0,
      "votes": 6
    },
    {
      "id": 125259,
      "title": "Dataset details",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 128560,
      "title": "Does submitting score always takes a long time?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 128069,
      "title": "Frame 150 from ALL Videos",
      "comment_count": 8,
      "views": 0,
      "votes": 6
    },
    {
      "id": 124348,
      "title": "Dataset with extracted faces from first frame",
      "comment_count": 12,
      "views": 0,
      "votes": 11
    },
    {
      "id": 128223,
      "title": "non-standard dataset !!!",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 128319,
      "title": "Using Data beyond competition",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 127968,
      "title": "The Size of Private Test Set",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 127621,
      "title": "Submission Scoring error",
      "comment_count": 5,
      "views": 0,
      "votes": 3
    },
    {
      "id": 127877,
      "title": "How do you deal with such large amount of data?",
      "comment_count": 3,
      "views": 0,
      "votes": 0
    },
    {
      "id": 127832,
      "title": "AWS EC2 and SageMaker; which should I choose?",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 127837,
      "title": "Kernal crashing everytime I detect face using cv2 haarcascade",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 124383,
      "title": "Exposing DeepFake Videos By Detecting Face Warping Artifacts",
      "comment_count": 4,
      "views": 0,
      "votes": 5
    },
    {
      "id": 127504,
      "title": "Is this Competition for only Rich in Resources Kagglers",
      "comment_count": 14,
      "views": 0,
      "votes": null
    },
    {
      "id": 127997,
      "title": "How to load the dataset into Azure databricks",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 127912,
      "title": "How to get OpenCV GPU locally?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 127336,
      "title": "SAVI looks for inconsistencies in video to detect fakes",
      "comment_count": 6,
      "views": 0,
      "votes": 11
    },
    {
      "id": 127910,
      "title": "Leaderboard score based on test or train log loss?",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 127503,
      "title": "Can we PreProcess full data",
      "comment_count": 7,
      "views": 0,
      "votes": null
    },
    {
      "id": 127453,
      "title": "\"Media Forensics and DeepFakes: an overview\" - Great Paper",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 127683,
      "title": "Training Data #18 Has Mismatch",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 125044,
      "title": "Let's talk about Preprocessing",
      "comment_count": 46,
      "views": 0,
      "votes": 14
    },
    {
      "id": 127669,
      "title": "per image standardization rather than mean standardization?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 122276,
      "title": "Video Frame Processing Libraries in Offline Kaggle Docker Kernel",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 127536,
      "title": "Right Approach for Classification",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 127261,
      "title": "Uncertain scoring?",
      "comment_count": 6,
      "views": 0,
      "votes": 0
    },
    {
      "id": 126468,
      "title": "Watch your face!  🖼️👩‍🎨",
      "comment_count": 13,
      "views": 0,
      "votes": 15
    },
    {
      "id": 127353,
      "title": "AWS Promotional Credit Request (20200123) 2nd request deadline",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 127020,
      "title": "What should be input output path",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 121343,
      "title": "Asking for clarification on the code competition rules",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 127112,
      "title": "Trouble running a CPU-only kernel (unexplained timeout)",
      "comment_count": 3,
      "views": 0,
      "votes": 0
    },
    {
      "id": 126896,
      "title": "Kaggle Bug?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 127196,
      "title": "Anyone looking to create a new team? Please let me know! I am doing my masters in data science and looking to learn out of this exercise!",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 126286,
      "title": "How Deepfake Technology Can Become More Dangerous Than a Nuclear Weapon",
      "comment_count": 6,
      "views": 0,
      "votes": 10
    },
    {
      "id": 124256,
      "title": "Metadata update complete, rescore will begin on Monday",
      "comment_count": 32,
      "views": 0,
      "votes": 35
    },
    {
      "id": 126915,
      "title": "Getting Pretrained Model",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 126453,
      "title": "Frame dimensions for optimum performance",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 122785,
      "title": "Finding the fake audio",
      "comment_count": 15,
      "views": 0,
      "votes": 58
    },
    {
      "id": 127007,
      "title": "When we import VGG from Keras...",
      "comment_count": 1,
      "views": 0,
      "votes": 0
    },
    {
      "id": 125842,
      "title": "(Research question) Is there deepfake image segmentation paper?",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 126524,
      "title": "class balance on public data set",
      "comment_count": 0,
      "views": 0,
      "votes": 8
    },
    {
      "id": 125803,
      "title": "Kernel Ran Successfully but Failed at Submission",
      "comment_count": 7,
      "views": 0,
      "votes": 5
    },
    {
      "id": 126904,
      "title": "Size of public test set? (submission error)",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 126573,
      "title": "Can I access the /kaggle/working folder?",
      "comment_count": 10,
      "views": 0,
      "votes": 0
    },
    {
      "id": 126205,
      "title": "Fast and reliable face detection with RetinaFace",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 126554,
      "title": "Using locally trained model",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 126855,
      "title": "Rules concerning model checkpoint files",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 121262,
      "title": "download the full dataset",
      "comment_count": 10,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121240,
      "title": "A question about external data for the notebook",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 126786,
      "title": "Submission - Using external models",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 126668,
      "title": "How to crop face when I submit using kaggle notebook?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 124702,
      "title": "Submission CSV Not Found",
      "comment_count": 15,
      "views": 0,
      "votes": 8
    },
    {
      "id": 126211,
      "title": "cv2.VideoCapture() returns false for some videos",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 125206,
      "title": "Why so close?",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121982,
      "title": "Visual Comparison, are faces enough? videos + code example",
      "comment_count": 3,
      "views": 0,
      "votes": 24
    },
    {
      "id": 121694,
      "title": "REAL/FAKE is not enough, we need to know if its video/audio/both",
      "comment_count": 13,
      "views": 0,
      "votes": 14
    },
    {
      "id": 126646,
      "title": "Permission denied error in subprocess",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 125748,
      "title": "Has anybody successfully used ffmpeg on the Public Test Set?",
      "comment_count": 8,
      "views": 0,
      "votes": 6
    },
    {
      "id": 126496,
      "title": "Access to full dataset data via GCP (and other cloud providers)?",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 126540,
      "title": "How easy is it to make a deepfake video? | CNBC Reports",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 126334,
      "title": "Can we use the face recognition library?",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 126345,
      "title": "Face Only?",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 124182,
      "title": "How to use facenet-pytorch to crop faces from a frame?",
      "comment_count": 12,
      "views": 0,
      "votes": 7
    },
    {
      "id": 122113,
      "title": "IRB-HSR Documentation for Dataset",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 125637,
      "title": "kernel's GPU ON during public score?",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 126061,
      "title": "Question about customizing notebook environment",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 121484,
      "title": "Submission Issues",
      "comment_count": 27,
      "views": 0,
      "votes": 11
    },
    {
      "id": 125810,
      "title": "Kernel Commit",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 125827,
      "title": "Any way to download the training files into gcloud storage?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 125910,
      "title": "Submission Error",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 121483,
      "title": "Files in metadata.json but not in the training set",
      "comment_count": 8,
      "views": 0,
      "votes": 16
    },
    {
      "id": 121418,
      "title": "Reality Defender  - plugin for detecting fake videos",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 125420,
      "title": "Debugging the model",
      "comment_count": 8,
      "views": 0,
      "votes": 13
    },
    {
      "id": 122842,
      "title": "AWS GPU instances",
      "comment_count": 10,
      "views": 0,
      "votes": 8
    },
    {
      "id": 125590,
      "title": "Fix: Submission Scoring Error",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 125512,
      "title": "Can I train model on my local server and submit it?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 122286,
      "title": "What is the purpose of \"test_videos\"?",
      "comment_count": 9,
      "views": 0,
      "votes": 4
    },
    {
      "id": 125498,
      "title": "Accuracy scores",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 123444,
      "title": "Best Score without leak so far?",
      "comment_count": 42,
      "views": 0,
      "votes": 24
    },
    {
      "id": 122763,
      "title": "Files missing in train_sample_videos ?",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 125463,
      "title": "Compensating for head motion",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 125165,
      "title": "Interesting videos?",
      "comment_count": 1,
      "views": 0,
      "votes": 9
    },
    {
      "id": 121523,
      "title": "Requesting a GPU Face detection package : MTCNN",
      "comment_count": 9,
      "views": 0,
      "votes": 16
    },
    {
      "id": 125219,
      "title": "Submissions without any ML allowed?",
      "comment_count": 8,
      "views": 0,
      "votes": null
    },
    {
      "id": 125117,
      "title": "internet access !!!",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 125242,
      "title": "Original Files",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 125190,
      "title": "Are we allow to upload our trained models?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 125216,
      "title": "Question about eligibility for this challange",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 124923,
      "title": "Did you try the methods based on rPPG?",
      "comment_count": 3,
      "views": 0,
      "votes": 3
    },
    {
      "id": 124110,
      "title": "Frame by Frame Or Stream?",
      "comment_count": 12,
      "views": 0,
      "votes": 6
    },
    {
      "id": 125012,
      "title": "Does DFDC do the opposite of its objective?",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 124695,
      "title": "Missing Videos in Parts 18 and 35",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 124942,
      "title": "Metadata in subfolders?",
      "comment_count": 8,
      "views": 0,
      "votes": null
    },
    {
      "id": 124099,
      "title": "Reading a specific frame in a training video using OpenCv in Python",
      "comment_count": 7,
      "views": 0,
      "votes": 12
    },
    {
      "id": 124867,
      "title": "fake video with no fake?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 124717,
      "title": "Submission Scoring Error",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 124453,
      "title": "Notebook Exceeded Allowed Compute",
      "comment_count": 11,
      "views": 0,
      "votes": 1
    },
    {
      "id": 124933,
      "title": "How to find a proper pair of threshold and assignment when post processing?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 124694,
      "title": "A lot of these deepfakes are really bad",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 122841,
      "title": "Can we get a hint on submission errors?",
      "comment_count": 14,
      "views": 0,
      "votes": 34
    },
    {
      "id": 124590,
      "title": "Someone successfully submitted an audio classifier？get Submission Scoring Error when I extract audios on test set",
      "comment_count": 10,
      "views": 0,
      "votes": null
    },
    {
      "id": 124030,
      "title": "A Chocolate or a Shoe ? Deepfake, fake or Chocolate ",
      "comment_count": 9,
      "views": 0,
      "votes": 2
    },
    {
      "id": 124373,
      "title": "Beginner question: Upload data on AWS",
      "comment_count": 8,
      "views": 0,
      "votes": 6
    },
    {
      "id": 124607,
      "title": "Submission 17.26938",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 124568,
      "title": "Comparing real and fake faces side by side.",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 122141,
      "title": "Submission Error...",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 124535,
      "title": "Can an input have both audio and video altered or are they only done so exclusively?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 124130,
      "title": "StyleGAN2 - Dec 2019 paper",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 123260,
      "title": "Submission allways scores 0.69314",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 124247,
      "title": "MTCNN and GPU?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 124294,
      "title": "is it okay to use other kernel dataset?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 123733,
      "title": "I have made a nightmare",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 123875,
      "title": "What is your input: Image or Video?",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 122936,
      "title": "Effective Train Size",
      "comment_count": 12,
      "views": 0,
      "votes": 11
    },
    {
      "id": 122018,
      "title": "Anyone created Image dataset by sampling?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121443,
      "title": "Where can I find the metadata.json for file for all training videos (approximately 100k)?",
      "comment_count": 5,
      "views": 0,
      "votes": 4
    },
    {
      "id": 123863,
      "title": "S3 or EBS ?",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 123746,
      "title": "Video Tools Datasets",
      "comment_count": 1,
      "views": 0,
      "votes": 14
    },
    {
      "id": 123981,
      "title": "Liveness detection ",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 121250,
      "title": "You are limited to 9 hours of GPU compute time",
      "comment_count": 16,
      "views": 0,
      "votes": 8
    },
    {
      "id": 122634,
      "title": "Public test set size is 4000",
      "comment_count": 16,
      "views": 0,
      "votes": 30
    },
    {
      "id": 123948,
      "title": "Anyone using Caffe?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 121297,
      "title": "How can we know each video in the dataset is REAL or FAKE?",
      "comment_count": 10,
      "views": 0,
      "votes": 2
    },
    {
      "id": 123381,
      "title": "Another data leakage (threads and lookahead_threads)",
      "comment_count": 3,
      "views": 0,
      "votes": 6
    },
    {
      "id": 123771,
      "title": "Can't my kernel only output my submission.csv to LB?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 122620,
      "title": "Starting Point for Research",
      "comment_count": 4,
      "views": 0,
      "votes": 10
    },
    {
      "id": 121177,
      "title": "Looking for a Team",
      "comment_count": 8,
      "views": 0,
      "votes": 2
    },
    {
      "id": 123522,
      "title": "Query about Kernel Only Part.",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 122282,
      "title": "submission score is wrong ? ",
      "comment_count": 13,
      "views": 0,
      "votes": 4
    },
    {
      "id": 123424,
      "title": "Usage of older versions of tensorflow (1.14) without turning on the internet",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 121908,
      "title": "Where is the metadata.json files for training data?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 123486,
      "title": "Data Distribution in Public LB",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 122587,
      "title": "Easy Way to Face Alignment",
      "comment_count": 4,
      "views": 0,
      "votes": 9
    },
    {
      "id": 123483,
      "title": "what top scorer did",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 123109,
      "title": "Getting audio without external packages?",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 122991,
      "title": "Can anyone upload the metadata.json for all videos.",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 121330,
      "title": "16000 seconds just to load test set with OpenCV...",
      "comment_count": 12,
      "views": 0,
      "votes": 18
    },
    {
      "id": 121384,
      "title": "Can we get a definition of \"Deepfake\"?",
      "comment_count": 7,
      "views": 0,
      "votes": 1
    },
    {
      "id": 123209,
      "title": "Probability transformation in imbalanced datasets",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 123341,
      "title": "Deep Fake frames",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 123034,
      "title": "Is there any way to upgrade the package version or downgrade?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 122706,
      "title": "Beginner Question about the rules",
      "comment_count": 8,
      "views": 0,
      "votes": 1
    },
    {
      "id": 122837,
      "title": "Deepfakes in the cloud",
      "comment_count": 0,
      "views": 0,
      "votes": 10
    },
    {
      "id": 122854,
      "title": "\"Notebook Exceeded Allowed Compute\"",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 122887,
      "title": "Video length",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 123066,
      "title": "Server Error？",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 123026,
      "title": "Submission file question on loss",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 122856,
      "title": "CAN~T SUBMISSION !!! HELP",
      "comment_count": 4,
      "views": 0,
      "votes": 0
    },
    {
      "id": 122603,
      "title": "Is actually FaceForensics++ working well?",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 122736,
      "title": "Notebook timeout",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 122810,
      "title": "Is the 9hr on P100 applicable on Private Test Set as well?",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 122919,
      "title": "Erroneous submission => submission lost",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 122013,
      "title": "Public Test Set videos are not the same as Training Set",
      "comment_count": 0,
      "views": 0,
      "votes": 52
    },
    {
      "id": 121879,
      "title": "Real Vs Fake Audio Mel Frequency",
      "comment_count": 4,
      "views": 0,
      "votes": 14
    },
    {
      "id": 122702,
      "title": "Haar Cascade resources for face detection",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 122759,
      "title": "How to Download The Data via CLI?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121268,
      "title": "Fake audio detection could also help",
      "comment_count": 6,
      "views": 0,
      "votes": 10
    },
    {
      "id": 121854,
      "title": "Some Ideas/Thoughts About the Pipeline ",
      "comment_count": 6,
      "views": 0,
      "votes": 12
    },
    {
      "id": 121832,
      "title": "Ouch that data cap",
      "comment_count": 11,
      "views": 0,
      "votes": 2
    },
    {
      "id": 122436,
      "title": "Noob question about coding requirements: external packages",
      "comment_count": 3,
      "views": 0,
      "votes": 4
    },
    {
      "id": 121226,
      "title": "GCP Credits?",
      "comment_count": 10,
      "views": 0,
      "votes": 19
    },
    {
      "id": 121279,
      "title": "To the People worried about data size and hardware requirements",
      "comment_count": 26,
      "views": 0,
      "votes": 19
    },
    {
      "id": 121462,
      "title": "thousands in credits form AWS",
      "comment_count": 9,
      "views": 0,
      "votes": 19
    },
    {
      "id": 121735,
      "title": "Adversarial Meme :D ",
      "comment_count": 5,
      "views": 0,
      "votes": 14
    },
    {
      "id": 121228,
      "title": "Tips for participants with no hardware?",
      "comment_count": 7,
      "views": 0,
      "votes": 19
    },
    {
      "id": 121847,
      "title": "Any successful submission so far? (i.e. better than 0.69...)",
      "comment_count": 13,
      "views": 0,
      "votes": 8
    },
    {
      "id": 122122,
      "title": "how to import FaceswapGANModel",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 122359,
      "title": "Good validation set?",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 122316,
      "title": "Can anybody share the md5sum of dfdc_train_all.zip",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 121259,
      "title": "But where is the hash list?",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121399,
      "title": "Libraries available offline evaluation in kaggle environment",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 121687,
      "title": "All pre-trained models at one place",
      "comment_count": 5,
      "views": 0,
      "votes": 13
    },
    {
      "id": 122312,
      "title": "What’s the point of providing public test set?",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 122280,
      "title": "My submission has taken over 2 hours and has still not finished",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121461,
      "title": "One Face Per Frame?",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121989,
      "title": "Interesting Video Tutorials About Deepfakes",
      "comment_count": 1,
      "views": 0,
      "votes": 9
    },
    {
      "id": 122258,
      "title": "Estimates for best possible solutions error",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 122140,
      "title": "average hardware for this competition + optimization of data to run on at least a 1070",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 121791,
      "title": "When the facedetector is too good",
      "comment_count": 1,
      "views": 0,
      "votes": 7
    },
    {
      "id": 121373,
      "title": "Docker ffmpeg request",
      "comment_count": 5,
      "views": 0,
      "votes": 12
    },
    {
      "id": 121931,
      "title": "can not access '../input/deepfake-detection-challenge/' !!",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 122010,
      "title": "Submission CSV Not Found?Submission Scoring Error?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 121877,
      "title": "I install an package Matlab runtime but happen an error",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 121497,
      "title": "Any effective deep learning based face detection APIs or usable Github code?",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121315,
      "title": "1 Million Dollars ?",
      "comment_count": 13,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121707,
      "title": "Anyone else running into \"Notebook Exceeded Allowed Compute\"",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121717,
      "title": "Suggestion: Competition hosts sponsoring GPU hours",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 121428,
      "title": "Is sound usefull?",
      "comment_count": 6,
      "views": 0,
      "votes": 7
    },
    {
      "id": 121458,
      "title": "Small dataset",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121213,
      "title": "Explore image augmentations easily",
      "comment_count": 0,
      "views": 0,
      "votes": 12
    },
    {
      "id": 121656,
      "title": "First Folder compressed for Starters",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121524,
      "title": "Some stats",
      "comment_count": 3,
      "views": 0,
      "votes": 10
    },
    {
      "id": 121309,
      "title": "How does it calculate the score on test set?",
      "comment_count": 5,
      "views": 0,
      "votes": 3
    },
    {
      "id": 121403,
      "title": "Taking requests (and a funny picture)",
      "comment_count": 4,
      "views": 0,
      "votes": 8
    },
    {
      "id": 121533,
      "title": "Training on TPU inferencing on GPU",
      "comment_count": 7,
      "views": 0,
      "votes": 0
    },
    {
      "id": 121539,
      "title": "My notebook page is blank",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121387,
      "title": "Full House...  of mustaches",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 121420,
      "title": "Total dataset size after extraction?",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121355,
      "title": "Do we have more options aside cloud computing?",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121452,
      "title": "Helpful face detection/video tampering papers",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 121299,
      "title": "DeepFake Detector [Paper + video]",
      "comment_count": 1,
      "views": 0,
      "votes": 7
    },
    {
      "id": 121463,
      "title": "Need for speed",
      "comment_count": 8,
      "views": 0,
      "votes": -3
    },
    {
      "id": 121235,
      "title": "Huge amount of fakes?",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 121434,
      "title": "3d convolutions vs 2d convolutions",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121534,
      "title": "Unskew data",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121336,
      "title": "open and closed submissions?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 121257,
      "title": "Is it even possible to participate without a powerful hardware",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 121280,
      "title": "My blogpost about fake image detection",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121185,
      "title": "Question about the ELIGIBILITY of \"PAI steering committee organizations\"",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 121267,
      "title": "Placeholder Memes for the 6th Position Finish ",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 121466,
      "title": "Face swaps vs. reconstruction",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 121451,
      "title": "How can I submit my model that was trained by matlab ?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 121178,
      "title": "Hey early people! Anyone interested in teaming?",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 121192,
      "title": "Does an access to external computation resources play decisive role? ",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 121392,
      "title": "hardware specs",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 121346,
      "title": "Definition of AMLT",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121239,
      "title": "Accidental Admin Login",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121329,
      "title": "DeepFakes Explained",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 121258,
      "title": "Whats the best way to handle data?",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 121271,
      "title": "New to Kaggle, can I find teamates?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 121221,
      "title": "starter teamat",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 121220,
      "title": "For Starter",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    }
  ],
  "errors": []
}