| Benchmark | Model | Rank | Results |
|---|---|---|---|
| action-detection-on-j-hmdb | Faster-RCNN + two-stream I3D conv | #6 | Frame-mAP 0.5: 73.3Video-mAP 0.5: 78.6 |
| action-detection-on-ucf101-24 | Faster-RCNN + two-stream I3D conv | #6 | Frame-mAP 0.5: 76.3Video-mAP 0.5: 59.9 |
| action-recognition-in-videos-on-ava-v21 | S3D-G w/ ResNet RPN (Kinetics-400 pretraining( | #9 | mAP (Val): 22.0 |