Update: Added links to published papers.
\nThe CLEF 2021 conference will be all-virtual and free of charge. To participate, please register here. The LifeCLEF workshop will be help on Sep. 23rd, 11:30 GMT+3. Feel free to join!
\nThis competition was part of the 2021 LifeCLEF track and we asked for your working notes as contribution to the scientific field of bioacoustics. Eight teams followed our call and submitted papers summarizing their approach, models, post-processing and scores.
\nThank you to everyone who went the extra mile and submitted a working note!
\nWe promised we would review all submitted papers and pick the \"Best Working Note\" based on the three criteria \"Evaluation of work and contribution\", \"Originality and novelty\", and \"Readability and organization\". We asked reviewers to provide a score between 1 and 5 for each of the categories, and we then averaged all of these scores. Furthermore, we assigned three reviewers to each working note to get a representative assessment.
\nSo without further ado, here are the top three working notes of this year's edition:
\n1. Marcos V. Conde, Kumar Shubham, Prateek Agnihotri, Nitin D. Movva and Szilard Bessenyei: \"Weakly-Supervised Classification and Detection of Bird Sounds in the Wild. A BirdCLEF 2021 Solution\" (Team Ed and Satoru) with 4.1 average reviewer points.
\n2. Jean-Francois Puget: \"STFT Transformers for Bird Song Recognition\" (Team CPMP) with 4.0 average reviewer points.
\n3. Christof Henkel, Pascal Pfeiffer and Philipp Singer: \"Recognizing bird species in diverse soundscapes under weak supervision\" (Team new baseline) with 3.8 average reviewer points.
\nRanks 4-8 (alphabetically by first author name):
\nArunodhayan Sampathkumar and Danny Kowerko: \"TUC Media Computing at BIRDCLEF 2021:Noise augmentation strategies in bird sound classification in combination with DenseNets andResNets\" (Team Arunodhayan)
Gyanendra Das and Saksham Aggarwal: \"Bird-Species Audio Identification, Ensembling 1D + 2D Signals\" (Team Error_404)
Jan Schlüter: \"Learning to Monitor Birdcalls From Weakly-Labeled Focused Recordings\" (Team Jan Schlüter)
Maxim Shugaev, Naoya Tanahashi, Philip Dhingra and Urvish Patel: \"BirdCLEF 2021: building a birdcall segmentation model based on weak labels\" (Team Just do it)
Naoki Murakami, Hajime Tanaka and Masataka Nishimori: \"Birdcall Identification using CNN and Gradient Boosting Decision Trees with Weak and Noisy Supervision\" (Team Dr.北村の愉快な仲間たち)
Congratulations to Conde et al. for winning the best working note award of the 2021 BirdCLEF bird call identification competition!
\nAgain, thanks to everyone who submitted a working note. Please make sure to check them out and feel free to use them as reference for future competitions.
\nLet us know in the comments if you have any questions. Thanks to everyone who participated and made this the most successful BirdCLEF competition to date!
", "messages": [ { "id": "1387987", "postDate": "07/14/2021 15:00:03", "content": "Update: Added links to published papers.
\nThe CLEF 2021 conference will be all-virtual and free of charge. To participate, please register here. The LifeCLEF workshop will be help on Sep. 23rd, 11:30 GMT+3. Feel free to join!
\nThis competition was part of the 2021 LifeCLEF track and we asked for your working notes as contribution to the scientific field of bioacoustics. Eight teams followed our call and submitted papers summarizing their approach, models, post-processing and scores.
\nThank you to everyone who went the extra mile and submitted a working note!
\nWe promised we would review all submitted papers and pick the \"Best Working Note\" based on the three criteria \"Evaluation of work and contribution\", \"Originality and novelty\", and \"Readability and organization\". We asked reviewers to provide a score between 1 and 5 for each of the categories, and we then averaged all of these scores. Furthermore, we assigned three reviewers to each working note to get a representative assessment.
\nSo without further ado, here are the top three working notes of this year's edition:
\n1. Marcos V. Conde, Kumar Shubham, Prateek Agnihotri, Nitin D. Movva and Szilard Bessenyei: \"Weakly-Supervised Classification and Detection of Bird Sounds in the Wild. A BirdCLEF 2021 Solution\" (Team Ed and Satoru) with 4.1 average reviewer points.
\n2. Jean-Francois Puget: \"STFT Transformers for Bird Song Recognition\" (Team CPMP) with 4.0 average reviewer points.
\n3. Christof Henkel, Pascal Pfeiffer and Philipp Singer: \"Recognizing bird species in diverse soundscapes under weak supervision\" (Team new baseline) with 3.8 average reviewer points.
\nRanks 4-8 (alphabetically by first author name):
\nArunodhayan Sampathkumar and Danny Kowerko: \"TUC Media Computing at BIRDCLEF 2021:Noise augmentation strategies in bird sound classification in combination with DenseNets andResNets\" (Team Arunodhayan)
Gyanendra Das and Saksham Aggarwal: \"Bird-Species Audio Identification, Ensembling 1D + 2D Signals\" (Team Error_404)
Jan Schlüter: \"Learning to Monitor Birdcalls From Weakly-Labeled Focused Recordings\" (Team Jan Schlüter)
Maxim Shugaev, Naoya Tanahashi, Philip Dhingra and Urvish Patel: \"BirdCLEF 2021: building a birdcall segmentation model based on weak labels\" (Team Just do it)
Naoki Murakami, Hajime Tanaka and Masataka Nishimori: \"Birdcall Identification using CNN and Gradient Boosting Decision Trees with Weak and Noisy Supervision\" (Team Dr.北村の愉快な仲間たち)
Congratulations to Conde et al. for winning the best working note award of the 2021 BirdCLEF bird call identification competition!
\nAgain, thanks to everyone who submitted a working note. Please make sure to check them out and feel free to use them as reference for future competitions.
\nLet us know in the comments if you have any questions. Thanks to everyone who participated and made this the most successful BirdCLEF competition to date!
", "rawMarkdown": "**Update:** Added links to published papers. \n\n**The CLEF 2021 conference will be all-virtual and free of charge. To participate, please register [here](https://docs.google.com/forms/d/e/1FAIpQLSdn45KhRPwzolT1sRv35AUXU-P2afVaclrH2HJ44SxDfhVAFw/viewform). The [LifeCLEF workshop](https://www.imageclef.org/LifeCLEF2021) will be help on Sep. 23rd, 11:30 GMT+3. Feel free to join!**\n\nThis competition was part of the 2021 LifeCLEF track and we asked for your working notes as contribution to the scientific field of bioacoustics. Eight teams followed our call and submitted papers summarizing their approach, models, post-processing and scores.\n\nThank you to everyone who went the extra mile and submitted a working note!\n\nWe promised we would review all submitted papers and pick the \"Best Working Note\" based on the three criteria \"Evaluation of work and contribution\", \"Originality and novelty\", and \"Readability and organization\". We asked reviewers to provide a score between 1 and 5 for each of the categories, and we then averaged all of these scores. Furthermore, we assigned three reviewers to each working note to get a representative assessment.\n\nSo without further ado, here are the top three working notes of this year's edition:\n\n**1.** [Marcos V. Conde, Kumar Shubham, Prateek Agnihotri, Nitin D. Movva and Szilard Bessenyei: \"*Weakly-Supervised Classification and Detection of Bird Sounds in the Wild. A BirdCLEF 2021 Solution](http://ceur-ws.org/Vol-2936/paper-131.pdf)*\" (Team Ed and Satoru) with **4.1 average reviewer points**.\n\n**2.** [Jean-Francois Puget: \"*STFT Transformers for Bird Song Recognition](http://ceur-ws.org/Vol-2936/paper-137.pdf)*\" (Team CPMP) with **4.0 average reviewer points**.\n\n**3.** [Christof Henkel, Pascal Pfeiffer and Philipp Singer: \"*Recognizing bird species in diverse soundscapes under weak supervision](http://ceur-ws.org/Vol-2936/paper-134.pdf)*\" (Team new baseline) with **3.8 average reviewer points.**\n\nRanks 4-8 (alphabetically by first author name):\n\n- [Arunodhayan Sampathkumar and Danny Kowerko: \"*TUC Media Computing at BIRDCLEF 2021:Noise augmentation strategies in bird sound classification in combination with DenseNets andResNets](http://ceur-ws.org/Vol-2936/paper-138.pdf)*\" (Team Arunodhayan)\n\n- [Gyanendra Das and Saksham Aggarwal: \"*Bird-Species Audio Identification, Ensembling 1D + 2D Signals](http://ceur-ws.org/Vol-2936/paper-132.pdf)*\" (Team Error_404)\n\n- [Jan Schlüter: \"*Learning to Monitor Birdcalls From Weakly-Labeled Focused Recordings](http://ceur-ws.org/Vol-2936/paper-139.pdf)*\" (Team Jan Schlüter)\n\n- [Maxim Shugaev, Naoya Tanahashi, Philip Dhingra and Urvish Patel: \"*BirdCLEF 2021: building a birdcall segmentation model based on weak labels](http://ceur-ws.org/Vol-2936/paper-141.pdf)*\" (Team Just do it)\n\n- [Naoki Murakami, Hajime Tanaka and Masataka Nishimori: \"*Birdcall Identification using CNN and Gradient Boosting Decision Trees with Weak and Noisy Supervision](http://ceur-ws.org/Vol-2936/paper-136.pdf)*\" (Team Dr.北村の愉快な仲間たち)\n\nCongratulations to Conde et al. for winning the best working note award of the 2021 BirdCLEF bird call identification competition!\n\nAgain, thanks to everyone who submitted a working note. Please make sure to check them out and feel free to use them as reference for future competitions.\n\nLet us know in the comments if you have any questions. Thanks to everyone who participated and made this the most successful BirdCLEF competition to date!", "votes": null }, { "id": "1387998", "postDate": "07/14/2021 15:07:19", "content": "Wow, congrats to Conde at al. I am very happy to finish 2nd and see my colleague Christof take 3rd with his team mates. Well done all. I am sure I'll learn a lot from other teams papers.
", "rawMarkdown": "Wow, congrats to Conde at al. I am very happy to finish 2nd and see my colleague Christof take 3rd with his team mates. Well done all. I am sure I'll learn a lot from other teams papers.", "votes": null }, { "id": "1388002", "postDate": "07/14/2021 15:18:53", "content": "Thank you and congratulations to you too :)
", "rawMarkdown": "Thank you and congratulations to you too :)", "votes": null }, { "id": "1388187", "postDate": "07/14/2021 18:08:14", "content": "Congratulations to you as well @cpmpml and all the winning teams. Can't wait to read other papers !!
", "rawMarkdown": "Congratulations to you as well @cpmpml and all the winning teams. Can't wait to read other papers !!", "votes": null }, { "id": "1388520", "postDate": "07/15/2021 03:18:10", "content": "thanks, it is an honor to get your recognition :) congrats to all the winning teams and kagglers that decided to write down their approaches.
", "rawMarkdown": "thanks, it is an honor to get your recognition :) congrats to all the winning teams and kagglers that decided to write down their approaches.", "votes": null }, { "id": "1388799", "postDate": "07/15/2021 08:30:23", "content": "Now I am wondering who is who, i.e. how to match real names and Kaggle aliases…
", "rawMarkdown": "Now I am wondering who is who, i.e. how to match real names and Kaggle aliases...", "votes": null }, { "id": "1473673", "postDate": "08/15/2021 16:48:45", "content": "Congratulations @cpmpml and @nitindatta ! Very well deserved!
", "rawMarkdown": "Congratulations @cpmpml and @nitindatta ! Very well deserved!", "votes": null } ], "comments": [ { "id": 1387998, "author_name": "cpmpml", "author_url": "", "post_date": "07/14/2021 15:07:19", "content": "Wow, congrats to Conde at al. I am very happy to finish 2nd and see my colleague Christof take 3rd with his team mates. Well done all. I am sure I'll learn a lot from other teams papers.
", "votes": null, "replies": [ { "id": 1388002, "author_name": "nitindatta", "author_url": "", "post_date": "07/14/2021 15:18:53", "content": "Thank you and congratulations to you too :)
", "votes": null, "replies": [] }, { "id": 1388187, "author_name": "ks2019", "author_url": "", "post_date": "07/14/2021 18:08:14", "content": "Congratulations to you as well @cpmpml and all the winning teams. Can't wait to read other papers !!
", "votes": null, "replies": [] }, { "id": 1388520, "author_name": "jesucristo", "author_url": "", "post_date": "07/15/2021 03:18:10", "content": "thanks, it is an honor to get your recognition :) congrats to all the winning teams and kagglers that decided to write down their approaches.
", "votes": null, "replies": [] }, { "id": 1388799, "author_name": "cpmpml", "author_url": "", "post_date": "07/15/2021 08:30:23", "content": "Now I am wondering who is who, i.e. how to match real names and Kaggle aliases…
", "votes": null, "replies": [] }, { "id": 1473673, "author_name": "saurabhbagchi", "author_url": "", "post_date": "08/15/2021 16:48:45", "content": "Congratulations @cpmpml and @nitindatta ! Very well deserved!
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1387987": "**Update:** Added links to published papers. \n\n**The CLEF 2021 conference will be all-virtual and free of charge. To participate, please register [here](https://docs.google.com/forms/d/e/1FAIpQLSdn45KhRPwzolT1sRv35AUXU-P2afVaclrH2HJ44SxDfhVAFw/viewform). The [LifeCLEF workshop](https://www.imageclef.org/LifeCLEF2021) will be help on Sep. 23rd, 11:30 GMT+3. Feel free to join!**\n\nThis competition was part of the 2021 LifeCLEF track and we asked for your working notes as contribution to the scientific field of bioacoustics. Eight teams followed our call and submitted papers summarizing their approach, models, post-processing and scores.\n\nThank you to everyone who went the extra mile and submitted a working note!\n\nWe promised we would review all submitted papers and pick the \"Best Working Note\" based on the three criteria \"Evaluation of work and contribution\", \"Originality and novelty\", and \"Readability and organization\". We asked reviewers to provide a score between 1 and 5 for each of the categories, and we then averaged all of these scores. Furthermore, we assigned three reviewers to each working note to get a representative assessment.\n\nSo without further ado, here are the top three working notes of this year's edition:\n\n**1.** [Marcos V. Conde, Kumar Shubham, Prateek Agnihotri, Nitin D. Movva and Szilard Bessenyei: \"*Weakly-Supervised Classification and Detection of Bird Sounds in the Wild. A BirdCLEF 2021 Solution](http://ceur-ws.org/Vol-2936/paper-131.pdf)*\" (Team Ed and Satoru) with **4.1 average reviewer points**.\n\n**2.** [Jean-Francois Puget: \"*STFT Transformers for Bird Song Recognition](http://ceur-ws.org/Vol-2936/paper-137.pdf)*\" (Team CPMP) with **4.0 average reviewer points**.\n\n**3.** [Christof Henkel, Pascal Pfeiffer and Philipp Singer: \"*Recognizing bird species in diverse soundscapes under weak supervision](http://ceur-ws.org/Vol-2936/paper-134.pdf)*\" (Team new baseline) with **3.8 average reviewer points.**\n\nRanks 4-8 (alphabetically by first author name):\n\n- [Arunodhayan Sampathkumar and Danny Kowerko: \"*TUC Media Computing at BIRDCLEF 2021:Noise augmentation strategies in bird sound classification in combination with DenseNets andResNets](http://ceur-ws.org/Vol-2936/paper-138.pdf)*\" (Team Arunodhayan)\n\n- [Gyanendra Das and Saksham Aggarwal: \"*Bird-Species Audio Identification, Ensembling 1D + 2D Signals](http://ceur-ws.org/Vol-2936/paper-132.pdf)*\" (Team Error_404)\n\n- [Jan Schlüter: \"*Learning to Monitor Birdcalls From Weakly-Labeled Focused Recordings](http://ceur-ws.org/Vol-2936/paper-139.pdf)*\" (Team Jan Schlüter)\n\n- [Maxim Shugaev, Naoya Tanahashi, Philip Dhingra and Urvish Patel: \"*BirdCLEF 2021: building a birdcall segmentation model based on weak labels](http://ceur-ws.org/Vol-2936/paper-141.pdf)*\" (Team Just do it)\n\n- [Naoki Murakami, Hajime Tanaka and Masataka Nishimori: \"*Birdcall Identification using CNN and Gradient Boosting Decision Trees with Weak and Noisy Supervision](http://ceur-ws.org/Vol-2936/paper-136.pdf)*\" (Team Dr.北村の愉快な仲間たち)\n\nCongratulations to Conde et al. for winning the best working note award of the 2021 BirdCLEF bird call identification competition!\n\nAgain, thanks to everyone who submitted a working note. Please make sure to check them out and feel free to use them as reference for future competitions.\n\nLet us know in the comments if you have any questions. Thanks to everyone who participated and made this the most successful BirdCLEF competition to date!", "1387998": "Wow, congrats to Conde at al. I am very happy to finish 2nd and see my colleague Christof take 3rd with his team mates. Well done all. I am sure I'll learn a lot from other teams papers.", "1388002": "Thank you and congratulations to you too :)", "1388187": "Congratulations to you as well @cpmpml and all the winning teams. Can't wait to read other papers !!", "1388520": "thanks, it is an honor to get your recognition :) congrats to all the winning teams and kagglers that decided to write down their approaches.", "1388799": "Now I am wondering who is who, i.e. how to match real names and Kaggle aliases...", "1473673": "Congratulations @cpmpml and @nitindatta ! Very well deserved!" }, "source": "meta" }