You can deploy this model and start earning money today!

OpenSound/
EzAudioCaps

Views

No views yet

Prepare following dataset:

AudioSet: https://research.google.com/audioset/ AudioSet Strongly Labeled Subset: https://research.google.com/audioset/download_strong.html VggSound: https://www.robots.ox.ac.uk/~vgg/data/vggsound/ AudioCaps: https://audiocaps.github.io/

The audio file name format:

AudioSet data: {ytb_id}{start_second*1000}{end_second1000} AudioSet Strongly Labeled Subset: {ytb_id}_{start_second1000} (10s clip from the the start second) VggSound: {ytb_id}_{start_second} (10s… See the full description on the dataset page: https://huggingface.co/datasets/OpenSound/EzAudioCaps.