Swin-T (ImageNet-1k pretrain)

Reported on 8 benchmarks across 7 tasks · 1 paper

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

Computer Vision4 results

VideoonKinetics-400
Acc@1· 2021-06-24
78.8
best: 93.6 (OmniVec2)
Video Swin Transformer arXiv:2106.13230
VideoonKinetics-400
Acc@5· 2021-06-24
93.6
best: 98.9 (TubeViT-H (ImageNet-1k))
Video Swin Transformer arXiv:2106.13230
Object DetectiononTBBR
Average Recall@IoU:0.5-0.95
45.4
Instance SegmentationonTBBR
Average Recall@IoU:0.5-0.95
28

Methodology4 results

3DonTBBR
Average Recall@IoU:0.5-0.95
45.4
2D ClassificationonTBBR
Average Recall@IoU:0.5-0.95
45.4
2D Object DetectiononTBBR
Average Recall@IoU:0.5-0.95
45.4
16konTBBR
Average Recall@IoU:0.5-0.95
45.4