sed

[fuguanyu]

Hello, can this project be used for sound event detection? Input variable-length audio, and output chunk detection results at the frame level

[YuanGongND]

hi there,

Yes and no. We do not provide a function in this repo. But frame-wise embedding can be used as the basis for frame-level prediction. See https://arxiv.org/abs/2202.00874.

-Yuan