PMM(ResNet38, no saliency, no RW)

Reported on 6 benchmarks across 2 tasks · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Medical3 results

Semantic SegmentationonCOCO 2014 val
mIoU· 2021-08-30
36.7
best: 56.8 (DHR (Swin-L, Mask2Former))
Pseudo-mask Matters in Weakly-supervised Semantic Segmentation arXiv:2108.12995
Semantic SegmentationonPASCAL VOC 2012 val
Mean IoU· 2021-08-30
68.5
best: 83.4 (SemPLeS (Swin-L))
Pseudo-mask Matters in Weakly-supervised Semantic Segmentation arXiv:2108.12995
Semantic SegmentationonPASCAL VOC 2012 test
Mean IoU· 2021-08-30
69
best: 82.9 (SemPLeS (Swin-L))
Pseudo-mask Matters in Weakly-supervised Semantic Segmentation arXiv:2108.12995

Audio3 results

10-shot image generationonCOCO 2014 val
mIoU· 2021-08-30
36.7
best: 56.8 (DHR (Swin-L, Mask2Former))
Pseudo-mask Matters in Weakly-supervised Semantic Segmentation arXiv:2108.12995
10-shot image generationonPASCAL VOC 2012 val
Mean IoU· 2021-08-30
68.5
best: 83.4 (SemPLeS (Swin-L))
Pseudo-mask Matters in Weakly-supervised Semantic Segmentation arXiv:2108.12995
10-shot image generationonPASCAL VOC 2012 test
Mean IoU· 2021-08-30
69
best: 82.9 (SemPLeS (Swin-L))
Pseudo-mask Matters in Weakly-supervised Semantic Segmentation arXiv:2108.12995