{ "id": 87986, "title": "A Welcome Message by the Competition Organizers", "url": "/competitions/freesound-audio-tagging-2019/discussion/87986", "author_name": "Eduardo Fonseca", "post_date": "2019-04-05T00:01:12.823000", "votes": 4, "comment_count": 0, "views": 0, "content": "

Welcome to Freesound Audio Tagging 2019!

\n\n

We are Eduardo and Frederic from the Freesound team at MTG-UPF, and Manoj and Dan from the Sound Understanding team at Google AI, and we are excited to bring you this competition!

\n\n

This is a follow-up of last year's Freesound General-Purpose Audio Tagging Challenge.

\n\n

This time we are taking the challenge to the next level with multi-label audio tagging, twice the number of audio categories, and a noisier-than-ever training set. You are challenged to build models that can recognize 80 diverse categories of everyday sounds using a train set composed mainly of noisy labels.

\n\n

This competition is part of the DCASE 2019 Challenge, an academic competition featuring five tasks related to the computational analysis of sound events and scenes (we are Task 2). Results of the DCASE Challenge will be presented and discussed at the DCASE 2019 Workshop. We encourage you to submit a paper to the DCASE Workshop describing your system for this competition. You can find more information in the competition’s DCASE tab.

\n\n

Please feel free to ask your questions in this thread.

\n\n

Good luck!

\n\n

Eduardo Fonseca, Frederic Font (Freesound, MTG-UPF)
\nManoj Plakal, Dan Ellis (Sound Understanding, Google AI)

", "messages": [ { "id": 507604, "postDate": "2019-04-05T00:01:12.823Z", "content": "

Welcome to Freesound Audio Tagging 2019!

\n\n

We are Eduardo and Frederic from the Freesound team at MTG-UPF, and Manoj and Dan from the Sound Understanding team at Google AI, and we are excited to bring you this competition!

\n\n

This is a follow-up of last year's Freesound General-Purpose Audio Tagging Challenge.

\n\n

This time we are taking the challenge to the next level with multi-label audio tagging, twice the number of audio categories, and a noisier-than-ever training set. You are challenged to build models that can recognize 80 diverse categories of everyday sounds using a train set composed mainly of noisy labels.

\n\n

This competition is part of the DCASE 2019 Challenge, an academic competition featuring five tasks related to the computational analysis of sound events and scenes (we are Task 2). Results of the DCASE Challenge will be presented and discussed at the DCASE 2019 Workshop. We encourage you to submit a paper to the DCASE Workshop describing your system for this competition. You can find more information in the competition’s DCASE tab.

\n\n

Please feel free to ask your questions in this thread.

\n\n

Good luck!

\n\n

Eduardo Fonseca, Frederic Font (Freesound, MTG-UPF)
\nManoj Plakal, Dan Ellis (Sound Understanding, Google AI)

", "rawMarkdown": "Welcome to **Freesound Audio Tagging 2019**! \n\nWe are Eduardo and Frederic from the Freesound team at MTG-UPF, and Manoj and Dan from the Sound Understanding team at Google AI, and we are excited to bring you this competition!\n\nThis is a follow-up of last year's Freesound General-Purpose Audio Tagging Challenge.\n\nThis time we are taking the challenge to the next level with multi-label audio tagging, twice the number of audio categories, and a *noisier-than-ever* training set. You are challenged to build models that can recognize 80 diverse categories of everyday sounds using a train set composed mainly of noisy labels.\n\nThis competition is part of the DCASE 2019 Challenge, an academic competition featuring five tasks related to the computational analysis of sound events and scenes (we are Task 2). Results of the DCASE Challenge will be presented and discussed at the DCASE 2019 Workshop. We encourage you to submit a paper to the DCASE Workshop describing your system for this competition. You can find more information in the competition’s DCASE tab.\n\nPlease feel free to ask your questions in this thread. \n\nGood luck!\n\nEduardo Fonseca, Frederic Font (Freesound, MTG-UPF) \nManoj Plakal, Dan Ellis (Sound Understanding, Google AI)", "votes": 4 } ], "comments": [], "raw_markdown_by_id": { "507604": "Welcome to **Freesound Audio Tagging 2019**! \n\nWe are Eduardo and Frederic from the Freesound team at MTG-UPF, and Manoj and Dan from the Sound Understanding team at Google AI, and we are excited to bring you this competition!\n\nThis is a follow-up of last year's Freesound General-Purpose Audio Tagging Challenge.\n\nThis time we are taking the challenge to the next level with multi-label audio tagging, twice the number of audio categories, and a *noisier-than-ever* training set. You are challenged to build models that can recognize 80 diverse categories of everyday sounds using a train set composed mainly of noisy labels.\n\nThis competition is part of the DCASE 2019 Challenge, an academic competition featuring five tasks related to the computational analysis of sound events and scenes (we are Task 2). Results of the DCASE Challenge will be presented and discussed at the DCASE 2019 Workshop. We encourage you to submit a paper to the DCASE Workshop describing your system for this competition. You can find more information in the competition’s DCASE tab.\n\nPlease feel free to ask your questions in this thread. \n\nGood luck!\n\nEduardo Fonseca, Frederic Font (Freesound, MTG-UPF) \nManoj Plakal, Dan Ellis (Sound Understanding, Google AI)" } }