{ "id": 412823, "title": "Resource you need From This Competition", "url": "/competitions/birdclef-2023/discussion/412823", "author_name": "Gaju Ahmed", "post_date": "2023-05-25T12:05:51.953000", "votes": 0, "comment_count": 0, "views": 0, "content": "

Discussion in this competition:

\n

[LB: 0.80] Pretraining is All you Need
\nPytorch Lightning Baseline - LB 0.78 CV 0.831
\nThis competition is using a new Python metrics framework
\n[LB: 0.78] BirdCLEF 2023 Baseline
\nIdentifying Audio Duplications
\nBirdNET usage
\n32kHz ogg additional data
\nTips for BirdCLEF series
\nAudio Signal Processing for Machine Learning (Video Series)
\nFaster inference using concurrent ThreadPoolExecutor.

\n

Best Notebook in this Competition:

\n

BirdCLEF23: Pretraining is All you Need [Train]
\nBirdCLEF23: Pretraining is All you Need [Infer]
\nBirdCLEF23: EffNet + FSR + CutMixUp [Train]
\nEDA|🐦Visualizations + Audio Exploration 🔉
\n[EDA] 🕊️BirdCLEF ~ 📊Plotly vis | 🌍Map | 🔊Audio
\nBirdClef 2023: Pytorch Lightning-Inference
\nBirdClef 2023: Pytorch Lightning-Training w/ cMAP
\nFaster eb0_SED model inference
\nBirdCLEF23: EffNet + FSR + CutMixUp [Infer]
\n

\n

Best Solution of this Competition:

\n

1st place solution: Correct Data is All You Need
\n2nd place solution: SED + CNN with 7 models ensemble
\n4th Place Solution: Knowledge Distillation is all you need
\n6th place solution: BirdNET embedding + CNN
\n9th Place Solution: 7 CNN Models Ensemble
\n10th place solution with the help of ChatGPT
\n12th place solution: 8 CNN models ensemble with OpenVINO
\n20th place solution: SED + CNN ensemble using onnx
\n
\n

", "messages": [ { "id": 2273784, "postDate": "2023-05-25T12:05:51.953Z", "content": "

Discussion in this competition:

\n

[LB: 0.80] Pretraining is All you Need
\nPytorch Lightning Baseline - LB 0.78 CV 0.831
\nThis competition is using a new Python metrics framework
\n[LB: 0.78] BirdCLEF 2023 Baseline
\nIdentifying Audio Duplications
\nBirdNET usage
\n32kHz ogg additional data
\nTips for BirdCLEF series
\nAudio Signal Processing for Machine Learning (Video Series)
\nFaster inference using concurrent ThreadPoolExecutor.

\n

Best Notebook in this Competition:

\n

BirdCLEF23: Pretraining is All you Need [Train]
\nBirdCLEF23: Pretraining is All you Need [Infer]
\nBirdCLEF23: EffNet + FSR + CutMixUp [Train]
\nEDA|🐦Visualizations + Audio Exploration 🔉
\n[EDA] 🕊️BirdCLEF ~ 📊Plotly vis | 🌍Map | 🔊Audio
\nBirdClef 2023: Pytorch Lightning-Inference
\nBirdClef 2023: Pytorch Lightning-Training w/ cMAP
\nFaster eb0_SED model inference
\nBirdCLEF23: EffNet + FSR + CutMixUp [Infer]
\n

\n

Best Solution of this Competition:

\n

1st place solution: Correct Data is All You Need
\n2nd place solution: SED + CNN with 7 models ensemble
\n4th Place Solution: Knowledge Distillation is all you need
\n6th place solution: BirdNET embedding + CNN
\n9th Place Solution: 7 CNN Models Ensemble
\n10th place solution with the help of ChatGPT
\n12th place solution: 8 CNN models ensemble with OpenVINO
\n20th place solution: SED + CNN ensemble using onnx
\n
\n

", "rawMarkdown": "## Discussion in this competition: \n[LB: 0.80] Pretraining is All you Need\nPytorch Lightning Baseline - LB 0.78 CV 0.831\nThis competition is using a new Python metrics framework\n[LB: 0.78] BirdCLEF 2023 Baseline\nIdentifying Audio Duplications\nBirdNET usage\n32kHz ogg additional data\nTips for BirdCLEF series\nAudio Signal Processing for Machine Learning (Video Series)\nFaster inference using concurrent ThreadPoolExecutor.\n\n## Best Notebook in this Competition:\nBirdCLEF23: Pretraining is All you Need [Train]\nBirdCLEF23: Pretraining is All you Need [Infer]\nBirdCLEF23: EffNet + FSR + CutMixUp [Train]\nEDA|🐦Visualizations + Audio Exploration 🔉\n[EDA] 🕊️BirdCLEF ~ 📊Plotly vis | 🌍Map | 🔊Audio\nBirdClef 2023: Pytorch Lightning-Inference\nBirdClef 2023: Pytorch Lightning-Training w/ cMAP\nFaster eb0_SED model inference\nBirdCLEF23: EffNet + FSR + CutMixUp [Infer]\n\n\n## Best Solution of this Competition:\n1st place solution: Correct Data is All You Need\n2nd place solution: SED + CNN with 7 models ensemble\n4th Place Solution: Knowledge Distillation is all you need\n6th place solution: BirdNET embedding + CNN\n9th Place Solution: 7 CNN Models Ensemble\n10th place solution with the help of ChatGPT\n12th place solution: 8 CNN models ensemble with OpenVINO\n20th place solution: SED + CNN ensemble using onnx\n\n\n" } ], "comments": [], "raw_markdown_by_id": { "2273784": "## Discussion in this competition: \n[LB: 0.80] Pretraining is All you Need\nPytorch Lightning Baseline - LB 0.78 CV 0.831\nThis competition is using a new Python metrics framework\n[LB: 0.78] BirdCLEF 2023 Baseline\nIdentifying Audio Duplications\nBirdNET usage\n32kHz ogg additional data\nTips for BirdCLEF series\nAudio Signal Processing for Machine Learning (Video Series)\nFaster inference using concurrent ThreadPoolExecutor.\n\n## Best Notebook in this Competition:\nBirdCLEF23: Pretraining is All you Need [Train]\nBirdCLEF23: Pretraining is All you Need [Infer]\nBirdCLEF23: EffNet + FSR + CutMixUp [Train]\nEDA|🐦Visualizations + Audio Exploration 🔉\n[EDA] 🕊️BirdCLEF ~ 📊Plotly vis | 🌍Map | 🔊Audio\nBirdClef 2023: Pytorch Lightning-Inference\nBirdClef 2023: Pytorch Lightning-Training w/ cMAP\nFaster eb0_SED model inference\nBirdCLEF23: EffNet + FSR + CutMixUp [Infer]\n\n\n## Best Solution of this Competition:\n1st place solution: Correct Data is All You Need\n2nd place solution: SED + CNN with 7 models ensemble\n4th Place Solution: Knowledge Distillation is all you need\n6th place solution: BirdNET embedding + CNN\n9th Place Solution: 7 CNN Models Ensemble\n10th place solution with the help of ChatGPT\n12th place solution: 8 CNN models ensemble with OpenVINO\n20th place solution: SED + CNN ensemble using onnx\n\n\n" } }