Finding Action Tubes with a Sparse-to-Dense Framework
Open paper
Benchmark
Model
Rank
Results
action-detection-on-j-hmdb
DTS
–
Video-mAP 0.2: 76.1
Video-mAP 0.5: 74.3
action-detection-on-ucf101-24
DTS
–
Video-mAP 0.5: 54
Rank counts only results with a code link.