{
  "total": 261,
  "topics": [
    {
      "id": 240323,
      "title": "Thank you! A survey and info on article for Nature Methods",
      "comment_count": 18,
      "views": 0,
      "votes": 29
    },
    {
      "id": 237905,
      "title": "Submission selection reminder",
      "comment_count": 17,
      "views": 0,
      "votes": 5
    },
    {
      "id": 230751,
      "title": "Public data as Kaggle dataset released",
      "comment_count": 6,
      "views": 0,
      "votes": 22
    },
    {
      "id": 223645,
      "title": "AMA 4-5pm CET on March 8th [still monitoring]",
      "comment_count": 42,
      "views": 0,
      "votes": 29
    },
    {
      "id": 214517,
      "title": "Looking for a Team Megathread",
      "comment_count": 109,
      "views": 0,
      "votes": 16
    },
    {
      "id": 214616,
      "title": "Welcome to the 2nd HPA challenge!",
      "comment_count": 50,
      "views": 0,
      "votes": 79
    },
    {
      "id": 217021,
      "title": "Updates complete",
      "comment_count": 15,
      "views": 0,
      "votes": 7
    },
    {
      "id": 215428,
      "title": "Updates coming!",
      "comment_count": 9,
      "views": 0,
      "votes": 34
    },
    {
      "id": 214516,
      "title": "New to Machine Learning or to Kaggle? Check this out ~",
      "comment_count": 4,
      "views": 0,
      "votes": 8
    },
    {
      "id": 578610,
      "title": "Private test set available post-challenge?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 214734,
      "title": "Self-supervised pretraining for this competition",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 214827,
      "title": "BIOLOGY 101",
      "comment_count": 8,
      "views": 0,
      "votes": 31
    },
    {
      "id": 356766,
      "title": "Published Paper in Nature Methods",
      "comment_count": 1,
      "views": 0,
      "votes": 15
    },
    {
      "id": 238365,
      "title": "What a ride! Parts of 6th place solution",
      "comment_count": 15,
      "views": 0,
      "votes": 42
    },
    {
      "id": 215773,
      "title": "Segmentation Mask Dataset",
      "comment_count": 21,
      "views": 0,
      "votes": 76
    },
    {
      "id": 232396,
      "title": "Model-centric vs data-centric",
      "comment_count": 6,
      "views": 0,
      "votes": 22
    },
    {
      "id": 276750,
      "title": "Single-cell RNASeq Analysis",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 239001,
      "title": "Fair Cell Activation Network and Swin Transformer, the 1st place solution",
      "comment_count": 33,
      "views": 0,
      "votes": 221
    },
    {
      "id": 238645,
      "title": "HPA 2nd Place Solution [red.ai]",
      "comment_count": 29,
      "views": 0,
      "votes": 63
    },
    {
      "id": 238862,
      "title": "3rd place solution - MPWARE part",
      "comment_count": 14,
      "views": 0,
      "votes": 43
    },
    {
      "id": 238487,
      "title": "5th place solution - my part",
      "comment_count": 14,
      "views": 0,
      "votes": 35
    },
    {
      "id": 238507,
      "title": "7th place solution",
      "comment_count": 11,
      "views": 0,
      "votes": 29
    },
    {
      "id": 239071,
      "title": "4th Place Solution: MILIMED",
      "comment_count": 5,
      "views": 0,
      "votes": 15
    },
    {
      "id": 232299,
      "title": "Side effects of HPA competition",
      "comment_count": 18,
      "views": 0,
      "votes": 95
    },
    {
      "id": 238371,
      "title": "A simple approach and only 2 cell-wise models to get a silver [LB Private 0.436]",
      "comment_count": 6,
      "views": 0,
      "votes": 30
    },
    {
      "id": 214817,
      "title": "Machine Learning in Single Cell Biology | Resources, Papers and articles",
      "comment_count": 2,
      "views": 0,
      "votes": 8
    },
    {
      "id": 242253,
      "title": "Is a score of 0.56 enough ?",
      "comment_count": 7,
      "views": 0,
      "votes": 3
    },
    {
      "id": 229376,
      "title": "How to use GPU in creating submission",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 238527,
      "title": "simple idea of 5th, tito's part",
      "comment_count": 14,
      "views": 0,
      "votes": 49
    },
    {
      "id": 241770,
      "title": "How to create non-leaking folds",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 241637,
      "title": "32nd Place Solution",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 238729,
      "title": "Meta-summary of the Top 30",
      "comment_count": 10,
      "views": 0,
      "votes": 36
    },
    {
      "id": 241355,
      "title": "28th solution/ with self-supervised learning",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 239166,
      "title": "6th place Solution Summary(0.549) ",
      "comment_count": 1,
      "views": 0,
      "votes": 24
    },
    {
      "id": 238898,
      "title": "3rd place Dieter part",
      "comment_count": 7,
      "views": 0,
      "votes": 41
    },
    {
      "id": 239266,
      "title": "Congratulations Teammate Darek (thedrcat) For Promotion to Kaggle Competitons Master!!!",
      "comment_count": 4,
      "views": 0,
      "votes": 32
    },
    {
      "id": 238361,
      "title": "8th place solution",
      "comment_count": 20,
      "views": 0,
      "votes": 55
    },
    {
      "id": 238512,
      "title": "23rd Place Solution: You don't need heuristics or expensive GPU",
      "comment_count": 17,
      "views": 0,
      "votes": 35
    },
    {
      "id": 239827,
      "title": "Hybrid approach - ML and GOP (good old programming)",
      "comment_count": 2,
      "views": 0,
      "votes": 10
    },
    {
      "id": 239861,
      "title": "HPA 36th Place Solution",
      "comment_count": 2,
      "views": 0,
      "votes": 10
    },
    {
      "id": 238678,
      "title": "9th Place Solution",
      "comment_count": 5,
      "views": 0,
      "votes": 24
    },
    {
      "id": 239641,
      "title": "20th Place Solution - Two stage model",
      "comment_count": 0,
      "views": 0,
      "votes": 7
    },
    {
      "id": 238547,
      "title": "A Thank-You To The Kaggle Community And My Team",
      "comment_count": 3,
      "views": 0,
      "votes": 18
    },
    {
      "id": 238910,
      "title": "3rd place solution - ZFTurbo part",
      "comment_count": 6,
      "views": 0,
      "votes": 28
    },
    {
      "id": 238333,
      "title": "13th Place Solution Summary - CVSSP & forecom.ai (0.527 LB)",
      "comment_count": 4,
      "views": 0,
      "votes": 40
    },
    {
      "id": 238364,
      "title": "10th Place Solution",
      "comment_count": 4,
      "views": 0,
      "votes": 27
    },
    {
      "id": 239203,
      "title": "💥Congratulations to bestfitting new #1 in competitions! 🖖👍",
      "comment_count": 0,
      "views": 0,
      "votes": 12
    },
    {
      "id": 238624,
      "title": "41st Place Solution: Cell-based RoI Pooling + Transformer Encoder",
      "comment_count": 4,
      "views": 0,
      "votes": 11
    },
    {
      "id": 239048,
      "title": "31rd Place Solution: CAM only.",
      "comment_count": 2,
      "views": 0,
      "votes": 15
    },
    {
      "id": 238474,
      "title": "21st Place Solution: You don't need cell tiles",
      "comment_count": 12,
      "views": 0,
      "votes": 27
    },
    {
      "id": 239099,
      "title": "How to get a gold medal in 2 weeks?",
      "comment_count": 2,
      "views": 0,
      "votes": 9
    },
    {
      "id": 238401,
      "title": "24th place (private) solution - Single model approach",
      "comment_count": 14,
      "views": 0,
      "votes": 32
    },
    {
      "id": 239211,
      "title": "11st Place Solution Summary: Active learning for cell label mining and Ensemble of different models",
      "comment_count": 0,
      "views": 0,
      "votes": 9
    },
    {
      "id": 238343,
      "title": "12th place solution summary (0.530 LB)",
      "comment_count": 6,
      "views": 0,
      "votes": 28
    },
    {
      "id": 238552,
      "title": "16th Place Solution: CNN, Yolov5, and use image-wise model to predict the cell",
      "comment_count": 9,
      "views": 0,
      "votes": 24
    },
    {
      "id": 238387,
      "title": "Private 22nd Place Solution",
      "comment_count": 15,
      "views": 0,
      "votes": 31
    },
    {
      "id": 238340,
      "title": "Congrats to all winners! [NEW GrandMaster and Master]",
      "comment_count": 8,
      "views": 0,
      "votes": 30
    },
    {
      "id": 238577,
      "title": "What a Kaggle week!",
      "comment_count": 3,
      "views": 0,
      "votes": 11
    },
    {
      "id": 238439,
      "title": "27th place solution (0.483LB)",
      "comment_count": 8,
      "views": 0,
      "votes": 17
    },
    {
      "id": 238380,
      "title": "Private 0.458 with 3 single cell classifiers (37th place solution) - a data centric approach",
      "comment_count": 1,
      "views": 0,
      "votes": 21
    },
    {
      "id": 237836,
      "title": "Notebook Running is worring",
      "comment_count": 18,
      "views": 0,
      "votes": 22
    },
    {
      "id": 238098,
      "title": "Hardware used for this competition?",
      "comment_count": 30,
      "views": 0,
      "votes": 8
    },
    {
      "id": 238984,
      "title": "Mitotic Spindle dataset",
      "comment_count": 0,
      "views": 0,
      "votes": 8
    },
    {
      "id": 238334,
      "title": "Thanks to the Amazing Host Team!!!",
      "comment_count": 4,
      "views": 0,
      "votes": 24
    },
    {
      "id": 238385,
      "title": "18th place solution private 0.515 public 0.514",
      "comment_count": 2,
      "views": 0,
      "votes": 17
    },
    {
      "id": 239035,
      "title": "15th Place Solution: CAM & Semantic Segmentation approach",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 238341,
      "title": "Thanks to the Great Team ! ",
      "comment_count": 8,
      "views": 0,
      "votes": 18
    },
    {
      "id": 238828,
      "title": "16bit or 8bit images? Does it make a difference?",
      "comment_count": 6,
      "views": 0,
      "votes": 5
    },
    {
      "id": 238504,
      "title": "43rd : Positive-Unlabeled Learning based Solution",
      "comment_count": 3,
      "views": 0,
      "votes": 12
    },
    {
      "id": 238348,
      "title": "💥💥Human Protein Atlas - Solution Compilation 💥💥",
      "comment_count": 0,
      "views": 0,
      "votes": 21
    },
    {
      "id": 238373,
      "title": "46th place solution - Simple Image Level Multilabel Classifier",
      "comment_count": 5,
      "views": 0,
      "votes": 15
    },
    {
      "id": 237374,
      "title": "Potential for shake up",
      "comment_count": 27,
      "views": 0,
      "votes": 17
    },
    {
      "id": 238561,
      "title": "tricks for high public LB score?",
      "comment_count": 4,
      "views": 0,
      "votes": 6
    },
    {
      "id": 238344,
      "title": "Just 2 more minutes, in the bonus area",
      "comment_count": 2,
      "views": 0,
      "votes": 7
    },
    {
      "id": 237745,
      "title": "Exact same csvs give scores differing by 0.3+",
      "comment_count": 23,
      "views": 0,
      "votes": 2
    },
    {
      "id": 238469,
      "title": "Last moment submission",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 237772,
      "title": "Notebook Timeout",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 237419,
      "title": "How does final submission on Wednesday actually work?",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 237573,
      "title": "OMG! same codes, but different score!",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 236623,
      "title": "Using HPACellSeg in final submissions",
      "comment_count": 10,
      "views": 0,
      "votes": 6
    },
    {
      "id": 236897,
      "title": "How about TTA",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 236916,
      "title": "Best Accuracy Score ?",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 236541,
      "title": "Labels on the test set",
      "comment_count": 9,
      "views": 0,
      "votes": 2
    },
    {
      "id": 230940,
      "title": "[LB 0.354 to 0.424] Add multi-label classification to the original image(green) ",
      "comment_count": 39,
      "views": 0,
      "votes": 40
    },
    {
      "id": 235846,
      "title": "Creating Save & Run All Version warnings.warn(msg, SourceChangeWarning)",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 235367,
      "title": "Why are we only asked to use the green filter to predict when blue filter indicates the nucleus label?",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 234654,
      "title": "HPACellSegmentator: CUDA out of memory.",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 235113,
      "title": "Cells do not stack",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 229962,
      "title": "The problem with using GoogleColab(GoogleDrive)",
      "comment_count": 9,
      "views": 0,
      "votes": 2
    },
    {
      "id": 233475,
      "title": "[external dataset] 16bit 1024x1024 HPA exteranl training data with cell instance mask",
      "comment_count": 1,
      "views": 0,
      "votes": 11
    },
    {
      "id": 233610,
      "title": "Domain level doubt !!!",
      "comment_count": 6,
      "views": 0,
      "votes": 5
    },
    {
      "id": 234282,
      "title": "Pytorch tips – loading data with speed and save models with more valuable information.",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 231582,
      "title": "[dataset] Single Label Cell Level Images TFRecords",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 233594,
      "title": "Submissions always score ~0.1 (need help)",
      "comment_count": 9,
      "views": 0,
      "votes": 2
    },
    {
      "id": 233479,
      "title": "[data aug] Using Albumentations for 4 Channel Image",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 233421,
      "title": "No output files found",
      "comment_count": 9,
      "views": 0,
      "votes": 1
    },
    {
      "id": 233378,
      "title": "Number of images in private test set",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 223822,
      "title": "HPA Public Data 768x768 \"rare classes\" dataset",
      "comment_count": 1,
      "views": 0,
      "votes": 16
    },
    {
      "id": 232715,
      "title": "Does external dataset include train dataset of last competition ?",
      "comment_count": 1,
      "views": 0,
      "votes": 8
    },
    {
      "id": 233056,
      "title": "Are yolov5 allowed for this competition?",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 230216,
      "title": "EfficientNetV2: Smaller Models and Faster Training",
      "comment_count": 3,
      "views": 0,
      "votes": 13
    },
    {
      "id": 233192,
      "title": "Now only 1 GPU instance allowed?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 232259,
      "title": "Grandmaster ratio",
      "comment_count": 5,
      "views": 0,
      "votes": 10
    },
    {
      "id": 233048,
      "title": "TracIn — A Simple Method to Estimate Training Data Influence (Google AI -NeurIPS 2020 paper)",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 231332,
      "title": "Help needed to debug \"Submission Scoring Error\"",
      "comment_count": 18,
      "views": 0,
      "votes": 3
    },
    {
      "id": 232588,
      "title": "Swin-Transformer",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 232906,
      "title": "Using TFRecord for Easier Data Manipulation",
      "comment_count": 2,
      "views": 0,
      "votes": 0
    },
    {
      "id": 232544,
      "title": "1024x1024 resized dataset (32.62GB)",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 232360,
      "title": "Notebook Exceeded Allowed Compute or Notebook Running while submission",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 232098,
      "title": "An idea for the use of weakly supervised labels",
      "comment_count": 1,
      "views": 0,
      "votes": 5
    },
    {
      "id": 231920,
      "title": "Better loss, Worse LB",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 232334,
      "title": "Evaluation metrics understanding help.",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 231637,
      "title": "Submission error help",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 231690,
      "title": "Code competitions errors and debugging tips (link)",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 231817,
      "title": "GPU quota + long save",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 229839,
      "title": "[dataset] 16bit 1024x1024 HPA training data with cell instance mask",
      "comment_count": 2,
      "views": 0,
      "votes": 21
    },
    {
      "id": 229586,
      "title": "Finally Completed Single Label Cell Level Dataset",
      "comment_count": 13,
      "views": 0,
      "votes": 26
    },
    {
      "id": 231036,
      "title": "hpacellseg.cellsegmentator cannot be downloaded?",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 222094,
      "title": "Puzzle-CAM",
      "comment_count": 61,
      "views": 0,
      "votes": 22
    },
    {
      "id": 231373,
      "title": "[HELP] Unable to fetch GCS Path of Kaggle Datasets",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 231184,
      "title": "Can the Human Protein Atlas Image Classification dataset be used?",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 230802,
      "title": "Using Cellpose for cell segmentation",
      "comment_count": 0,
      "views": 0,
      "votes": 10
    },
    {
      "id": 231166,
      "title": "Please help! Submission time is too high.",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 230023,
      "title": "can a cell be removed if it doesn't have a nucleus?  ",
      "comment_count": 3,
      "views": 0,
      "votes": 5
    },
    {
      "id": 222132,
      "title": "Even Faster Cell Segmentation",
      "comment_count": 2,
      "views": 0,
      "votes": 15
    },
    {
      "id": 230090,
      "title": "HPA-Cell-Segmentation Models missing",
      "comment_count": 5,
      "views": 0,
      "votes": 3
    },
    {
      "id": 230669,
      "title": "Instance Segmentation from Semantic Segmentation?",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 221748,
      "title": "Question about HPA public data ",
      "comment_count": 10,
      "views": 0,
      "votes": 7
    },
    {
      "id": 230658,
      "title": "Albumentations transforms with fast.ai DataBlock API",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 229661,
      "title": "The labels of each image, the labels of each cell",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 226930,
      "title": "Need help with submission",
      "comment_count": 18,
      "views": 0,
      "votes": 5
    },
    {
      "id": 229428,
      "title": "Model Evaluate Failure: Pls help me",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 229451,
      "title": "HPA masks not detected",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 229284,
      "title": "Relabel train dataset",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 228636,
      "title": "Looking for a teamate to work with",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 226894,
      "title": "A word of advice to Colab + GCS users",
      "comment_count": 8,
      "views": 0,
      "votes": 14
    },
    {
      "id": 229253,
      "title": "Few questions",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 227062,
      "title": "Dataset too large for Google Colab",
      "comment_count": 10,
      "views": 0,
      "votes": 2
    },
    {
      "id": 227785,
      "title": "Questions about the eval metrics on self-made val dataset",
      "comment_count": 8,
      "views": 0,
      "votes": 5
    },
    {
      "id": 216321,
      "title": "Baseline - Mmdetection Instance Segmentation Model",
      "comment_count": 8,
      "views": 0,
      "votes": 36
    },
    {
      "id": 221550,
      "title": "Rapid experimentation with fastai [0.342 lb]",
      "comment_count": 28,
      "views": 0,
      "votes": 64
    },
    {
      "id": 228008,
      "title": "Kaggle is so hard",
      "comment_count": 10,
      "views": 0,
      "votes": 27
    },
    {
      "id": 228635,
      "title": "What are the known techniques that can help in mitigaiting class imbalance in multi-label classification setting?",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 228960,
      "title": "Submission completion notifications",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 227862,
      "title": "There is one protein for each image, and its \"id\" (pdb) is known to hosts, but not disclosed for competition, is it correct ? ",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 215833,
      "title": "A possible reason for the \"Submission Scoring Error\"",
      "comment_count": 8,
      "views": 0,
      "votes": 9
    },
    {
      "id": 215351,
      "title": "How to Speed Up HPA Cell Segmentation?",
      "comment_count": 6,
      "views": 0,
      "votes": 7
    },
    {
      "id": 228604,
      "title": "Help with submission",
      "comment_count": 10,
      "views": 0,
      "votes": 1
    },
    {
      "id": 222370,
      "title": "Submission Scoring Error",
      "comment_count": 5,
      "views": 0,
      "votes": 4
    },
    {
      "id": 227822,
      "title": "Duplicate Images in Training Data",
      "comment_count": 2,
      "views": 0,
      "votes": 9
    },
    {
      "id": 227893,
      "title": "What do we predict? ",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 228374,
      "title": "Question about masks encodings",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 224073,
      "title": "I really can't figure out how to work with this 150GB+ data",
      "comment_count": 11,
      "views": 0,
      "votes": 4
    },
    {
      "id": 228132,
      "title": "About the using of cellsegmentator and submitting.",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 228109,
      "title": "Link to nice info on cell organelles ",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 219211,
      "title": "Question about Single-Labeled Images and Negative Label",
      "comment_count": 6,
      "views": 0,
      "votes": 2
    },
    {
      "id": 225449,
      "title": "3 techniques that might help in this competition",
      "comment_count": 4,
      "views": 0,
      "votes": 22
    },
    {
      "id": 227701,
      "title": "Number of Train images",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 225759,
      "title": "Is it possible to as kaggle increase memory size?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 226336,
      "title": "What does \"references\" mean in data description?",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 226500,
      "title": "Memory crash with more than 15K images",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 223281,
      "title": "Does notebook running time limit include the scoring time?",
      "comment_count": 8,
      "views": 0,
      "votes": 4
    },
    {
      "id": 225471,
      "title": "submit error",
      "comment_count": 10,
      "views": 0,
      "votes": 2
    },
    {
      "id": 226241,
      "title": "Mitotic spindle is beautiful",
      "comment_count": 1,
      "views": 0,
      "votes": 9
    },
    {
      "id": 227242,
      "title": "Question regarding prediction string in submission",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 220748,
      "title": "Proteins in Negatives",
      "comment_count": 18,
      "views": 0,
      "votes": 24
    },
    {
      "id": 224456,
      "title": "Class Names in Chinese【中文类别名称】",
      "comment_count": 10,
      "views": 0,
      "votes": 20
    },
    {
      "id": 215986,
      "title": "Insights from the previous HPA competition (video)",
      "comment_count": 1,
      "views": 0,
      "votes": 18
    },
    {
      "id": 226369,
      "title": "Sub-challenge: Locate the centrosome using the microtubules",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 225871,
      "title": "About Image Augmentation ",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 226197,
      "title": "Issue with submission",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 223745,
      "title": "HPA masks not working in test data",
      "comment_count": 5,
      "views": 0,
      "votes": 3
    },
    {
      "id": 215469,
      "title": "How to submit a non-zero score?",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 223886,
      "title": "What does 'weakly' mean?",
      "comment_count": 5,
      "views": 0,
      "votes": 5
    },
    {
      "id": 225402,
      "title": "Submissions taking longer than usual?",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 216926,
      "title": "Single-Stage Submission Approaches Overview [0.354 LB]",
      "comment_count": 27,
      "views": 0,
      "votes": 46
    },
    {
      "id": 225473,
      "title": "Build \"Submission file\"",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 225007,
      "title": "Questions about labeling",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 224305,
      "title": "Pip install extra python packages with no internet enabled",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 223441,
      "title": "How big is each pixel?",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 223934,
      "title": "Training image's protein channel suffers from translational artifact",
      "comment_count": 1,
      "views": 0,
      "votes": 10
    },
    {
      "id": 222655,
      "title": "Precomputing masks",
      "comment_count": 15,
      "views": 0,
      "votes": 2
    },
    {
      "id": 224291,
      "title": "Problem statement clarification",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 221022,
      "title": "Question About Multi-Label Classification",
      "comment_count": 19,
      "views": 0,
      "votes": 10
    },
    {
      "id": 217238,
      "title": "[survey] Conducting lesson for this competition",
      "comment_count": 25,
      "views": 0,
      "votes": 133
    },
    {
      "id": 222300,
      "title": "Can external data improve the performance？",
      "comment_count": 4,
      "views": 0,
      "votes": 5
    },
    {
      "id": 224005,
      "title": "Keeping interactive session live for 40+ mins",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 217806,
      "title": "Duplicate Classes in 146 Labels",
      "comment_count": 5,
      "views": 0,
      "votes": 27
    },
    {
      "id": 223610,
      "title": "Some explanations about predicted submissions?",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 223691,
      "title": "Issue with HPA cell segmentator - help needed!",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 223789,
      "title": "Predicting ER and Microtubules only based on yellow and red channels",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 219305,
      "title": "(Potential) Issues Found in HPACellSegmentator",
      "comment_count": 15,
      "views": 0,
      "votes": 18
    },
    {
      "id": 219885,
      "title": "What does confidence exactly do?",
      "comment_count": 23,
      "views": 0,
      "votes": 6
    },
    {
      "id": 222722,
      "title": "HPA Segmentation tool showing error for image size 1728?",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 222643,
      "title": "Using GPU for training models",
      "comment_count": 17,
      "views": 0,
      "votes": 1
    },
    {
      "id": 223167,
      "title": "Can't make sense out of \"sample submission\" file",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 220201,
      "title": "encoding string in prediction",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 221628,
      "title": "Submission time",
      "comment_count": 7,
      "views": 0,
      "votes": 3
    },
    {
      "id": 219375,
      "title": "Complete detail of predictionstring of sample_submission.csv with example.",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 221181,
      "title": "Submission for nothing detected or no cells?",
      "comment_count": 6,
      "views": 0,
      "votes": 10
    },
    {
      "id": 216984,
      "title": "Cell Crop Tile Datasets (256x256)",
      "comment_count": 5,
      "views": 0,
      "votes": 14
    },
    {
      "id": 221099,
      "title": "Does removing borderline cells matter any more with mAP?",
      "comment_count": 6,
      "views": 0,
      "votes": 7
    },
    {
      "id": 221015,
      "title": "Visualizing Segmentation Masks Interactively",
      "comment_count": 4,
      "views": 0,
      "votes": 11
    },
    {
      "id": 221405,
      "title": "need clarification on this competition",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 221025,
      "title": "Silly Question",
      "comment_count": 9,
      "views": 0,
      "votes": 3
    },
    {
      "id": 221421,
      "title": "How to handle Large datasets",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 216430,
      "title": "Train Dataset Containing All Cell Masks as RLE Strings",
      "comment_count": 4,
      "views": 0,
      "votes": 13
    },
    {
      "id": 214518,
      "title": "The Previous Human Protein Atlas Competition",
      "comment_count": 11,
      "views": 0,
      "votes": 52
    },
    {
      "id": 220275,
      "title": "Cell-level labels",
      "comment_count": 13,
      "views": 0,
      "votes": 3
    },
    {
      "id": 221251,
      "title": "Is it possible to participate of this competition, using R and reticulate r package? Thankyou!",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 219258,
      "title": "Best strategy to eval/monitort training",
      "comment_count": 2,
      "views": 0,
      "votes": 11
    },
    {
      "id": 220509,
      "title": "Where can I find the cell segmentator class?",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 214659,
      "title": "How to do local evaluation?",
      "comment_count": 14,
      "views": 0,
      "votes": 9
    },
    {
      "id": 218198,
      "title": "Interesting Paper/Code on Weakly Supervised Instance Segmentation Using MIL Framework",
      "comment_count": 1,
      "views": 0,
      "votes": 8
    },
    {
      "id": 215971,
      "title": "Does anyone knows what the feature description is when I decode the tfrec file?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 214835,
      "title": "Some resources about weakly-supervised segmentation",
      "comment_count": 4,
      "views": 0,
      "votes": 60
    },
    {
      "id": 215736,
      "title": "So, greens are the cells which the labels are tagged to?",
      "comment_count": 12,
      "views": 0,
      "votes": 18
    },
    {
      "id": 220375,
      "title": "Notebook Suggestions",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 215675,
      "title": "Which Proteins Are We Detecting?",
      "comment_count": 3,
      "views": 0,
      "votes": 3
    },
    {
      "id": 218309,
      "title": "Instance segmentation and its metric comprehension",
      "comment_count": 6,
      "views": 0,
      "votes": 6
    },
    {
      "id": 218732,
      "title": "Green Images",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 217395,
      "title": "[macro f1 publicLB0.43, mAP publicLB0.34] weakly supervised segmentation",
      "comment_count": 7,
      "views": 0,
      "votes": 99
    },
    {
      "id": 219539,
      "title": "A collection of some bioinformatics related resources on kaggle",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 215141,
      "title": "What does the ground truth look like?",
      "comment_count": 25,
      "views": 0,
      "votes": 26
    },
    {
      "id": 215981,
      "title": "Checking my understanding",
      "comment_count": 8,
      "views": 0,
      "votes": 14
    },
    {
      "id": 217158,
      "title": "a Few Tips on mAP",
      "comment_count": 2,
      "views": 0,
      "votes": 32
    },
    {
      "id": 218056,
      "title": "Commit Error",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 217714,
      "title": "Questions about local validation and submission",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 214701,
      "title": "Research Paper For Cell Classification",
      "comment_count": 9,
      "views": 0,
      "votes": 16
    },
    {
      "id": 217272,
      "title": "How to evaluate models and avoid overfitting in this competition?",
      "comment_count": 7,
      "views": 0,
      "votes": 7
    },
    {
      "id": 216924,
      "title": "Questions and Speculations on How To Handle the Negative Class",
      "comment_count": 4,
      "views": 0,
      "votes": 11
    },
    {
      "id": 217686,
      "title": "Target of this competition",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 217268,
      "title": "Please tell me about the data order of TFRecords.",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 216958,
      "title": "The best way to load uint8 and uint16",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 217048,
      "title": "EmbedSeg - Embedding-based Instance Segmentation of Microscopy Images",
      "comment_count": 2,
      "views": 0,
      "votes": 7
    },
    {
      "id": 216263,
      "title": "Option to choose to exclude tfrecord or other formats",
      "comment_count": 8,
      "views": 0,
      "votes": 11
    },
    {
      "id": 216732,
      "title": "Question About Training Image Level Labels",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 217179,
      "title": "New Kaggler: Downloading data?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 215122,
      "title": "Interesting labeling procedure",
      "comment_count": 2,
      "views": 0,
      "votes": 18
    },
    {
      "id": 216253,
      "title": "The fascinating world of a human cell",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 215238,
      "title": "resized dataset(512x512, 768x768)",
      "comment_count": 4,
      "views": 0,
      "votes": 44
    },
    {
      "id": 214526,
      "title": "Previous Human Protein Atlas Competition Nature Methods Paper",
      "comment_count": 1,
      "views": 0,
      "votes": 10
    },
    {
      "id": 216177,
      "title": "Why Are The Blue Channel Images SO Much Smaller?",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 214530,
      "title": "Potentially useful Github links",
      "comment_count": 9,
      "views": 0,
      "votes": 30
    },
    {
      "id": 214863,
      "title": "Convert From RGBY to RGB",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 214844,
      "title": "Fast pipeline resources for TFRecords+TPUs+Pytorch",
      "comment_count": 5,
      "views": 0,
      "votes": 9
    },
    {
      "id": 216099,
      "title": "Resources for Image/Cell Segmentation",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 215377,
      "title": "Bottleneck Transformers for Visual Recognition",
      "comment_count": 4,
      "views": 0,
      "votes": 8
    },
    {
      "id": 215172,
      "title": "Using HPA Cell Segmentation Model (No Internet)",
      "comment_count": 0,
      "views": 0,
      "votes": 25
    },
    {
      "id": 215772,
      "title": "Could anyone explain me the logic behind PredictionString for Submission.",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 215785,
      "title": "What's your best single model score?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 214905,
      "title": "HPA public data: How to speed up downloading?",
      "comment_count": 4,
      "views": 0,
      "votes": 5
    },
    {
      "id": 215805,
      "title": "Use code for reading images",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 215149,
      "title": "Useful approaches for subcellular protein localization problem",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 214603,
      "title": "How to visualize 4 channels in a single image? ",
      "comment_count": 1,
      "views": 0,
      "votes": 7
    },
    {
      "id": 214781,
      "title": "pycocotools",
      "comment_count": 1,
      "views": 0,
      "votes": 5
    },
    {
      "id": 214527,
      "title": "A link to human protein atlas external data",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 214891,
      "title": "Papers on multi-label classification protein and predicting protein patterns",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 214925,
      "title": "Summary of Human Protein Atlas - 1st Competition (Jan 2019)",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 215121,
      "title": "How can I avoid downloading the data?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 214528,
      "title": "Weak Supervised Learning",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    }
  ],
  "errors": []
}