Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

AVSBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Audio-Visual SegmentationAVSBench S4 v1 (test)
MJ86.2
55
Audio-Visual SegmentationAVSBench MS3 (test)
Jaccard Index (IoU)65
38
Audio-Visual SegmentationAVSBench MS3 v1 (test)
Mean Jaccard67.6
37
Audio-Visual Semantic SegmentationAVSBench AVSS v1 (test)
MJ51.2
29
Sound Target SegmentationAVSBench-object MS3 1.0 (test)
mIoU59.2
23
Audio-Visual SegmentationAVSBench Object S4 (test)
Jaccard Index (J)85.5
21
Audio-Visual SegmentationAVSBench AVS-Objects-MS3
J & F Score75.1
21
Audio-Visual SegmentationAVSBench AVS-Objects-S4
J&F Score92.4
21
Audio-Visual SegmentationAVSBench S4 (test)
F89.8
21
Audio-Visual Semantic SegmentationAVSBench-Semantic (AVSS) (test)
Jaccard Index (J)48.5
13
Audio-Visual SegmentationAVSBench AVS-Semantic
J (Jaccard)49.7
13
Sound Source SegmentationAVSBench
mIoU36.37
10
Audio referred image groundingAVSBench (test)
cIoU79.82
10
Audio-Visual Semantic SegmentationAVSBench Semantic (test)
mIoU50.1
8
Audio-Visual SegmentationAVSBench MS3 setting (test)
MJ Score55.1
6
Audio-visual segmentationAVSBench Multi-source
mIoU63.5
5
Audio-visual segmentationAVSBench Single-source
mIoU89.4
5
Audio-Visual SegmentationAVSBench Single Source (test)
mIoU34.13
5
Audio-Visual SegmentationAVSBench S4
mIoU80.4
5
Showing 19 of 19 rows