Nonspoken sound recognition and identification

[angrave]

Some interesting starting points- "Audio captioning" https://github.com/cdjkim/audiocaps

https://tango-web.github.io/ https://github.com/declare-lab/tango