AVSBench icon indicating copy to clipboard operation
AVSBench copied to clipboard

Published 20 hours ago • verified publisher icon OpenNLPLab
→

Metadata

[ECCV 2022] Official implementation of the paper: Audio-Visual Segmentation

Results 3 AVSBench issues
Sort by recently updated
recently updated
newest added

Have you released your best pretrained model weights of TPAVI

1 comment

SitongGong avatar SitongGong

Is it right to use 'bg_idx = (masks.shape[1] - 1)' ?

Hi, Your work is very interesting and I have a question about your code. Refer to line 42 in avss/loss.py, ` bg_idx = (pred_masks.shape[1] - 1)` you define background idx...

liushenme avatar liushenme

Could you please release your inference code?

txchen-USTC avatar txchen-USTC

← Metadata

431
Stars
40
Forks
Watchers

Owner

verified publisher icon OpenNLPLab

Metadata

[ECCV 2022] Official implementation of the paper: Audio-Visual Segmentation

Back