Temporal Action Localization on ActivityNet 1.2 (val)
45.3mAP@IoU 0.5CDC
Evaluation Results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CDCSupervision=Full2020.03 | 45.3 | — | — | 23.8 | — | — | — | — | — | — | — | |
| P-MILfused=true, fusion_detail=Fusion of S-MIL and P-MIL results2023.05 | 44.2 | 26.1 | 5.3 | 26.5 | — | — | — | — | — | — | — | |
| CO2-NetSupervision=Weakly2021.07 | 43.3 | 26.3 | 5.2 | 26.4 | — | — | — | — | — | — | — | |
| CO2-Net2023.05 | 43.3 | 26.3 | 5.2 | 26.4 | — | — | — | — | — | — | — | |
| COLA2023.05 | 42.7 | 25.7 | 5.8 | 26.1 | — | — | — | — | — | — | — | |
| D2-Net2023.05 | 42.3 | 25.5 | 5.8 | 26 | — | — | — | — | — | — | — | |
| P-MILfused=false2023.05 | 42.2 | 25 | 4.9 | 25.5 | — | — | — | — | — | — | — | |
| AUMN2021.04 | 42 | 25 | 5.6 | 25.5 | — | — | — | — | — | — | — | |
| AUMN2023.05 | 42 | 25 | 5.6 | 25.5 | — | — | — | — | — | — | — | |
| UGCT2023.05 | 41.8 | 25.3 | 5.9 | 25.8 | — | — | — | — | — | — | — | |
| SSNSupervision=Full2019.11 | 41.3 | 27 | 6.1 | 26.6 | — | — | — | — | — | — | — | |
| SSNSupervision=Full2020.03 | 41.3 | 27 | 6.1 | 26.6 | 38.8 | 35.9 | 32.9 | 30.4 | 22.2 | 18.2 | 13.2 | |
| SSNSupervision=Full2020.03 | 41.3 | — | — | 28.3 | — | — | — | 30.4 | — | — | 13.2 | |
| SSN (2017)Supervision Level=Full, Action Counts=False2020.06 | 41.3 | 27 | 6.1 | 26.6 | — | — | — | — | — | — | — | |
| SSNSupervision=Fully-Supervised2020.03 | 41.3 | — | — | 26.6 | — | — | — | 30.4 | — | — | 13.2 | |
| SSNSupervision=Full2021.03 | 41.3 | 27 | 6.1 | 26.6 | — | — | — | — | — | — | — | |
| SSNSupervision=Fully2021.07 | 41.3 | 27 | 6.1 | 26.6 | — | — | — | — | — | — | — | |
| Uncertainty Modeling for WTALSupervision Level=Weak, Action Counts=False2020.06 | 41.2 | 25.6 | 6 | 25.9 | — | — | — | — | — | — | — | |
| UMSupervision=Weakly2021.07 | 41.2 | 25.6 | 6 | 25.9 | — | — | — | — | — | — | — | |
| DGAMSupervision=Weak2020.03 | 41 | 23.5 | 5.3 | 24.4 | 37.5 | 33.5 | 30.1 | 26.9 | 19.8 | 15.5 | 10.8 | |
| DGAM (2020)Supervision Level=Weak, Action Counts=False2020.06 | 41 | 23.5 | 5.3 | 24.4 | — | — | — | — | — | — | — | |
| DGAMSupervision=Weak2021.03 | 41 | 23.5 | 5.3 | 24.4 | — | — | — | — | — | — | — | |
| DGAM2021.04 | 41 | 23.5 | 5.3 | 24.4 | — | — | — | — | — | — | — | |
| DGAMSupervision=Weakly Supervised, Backbone=I3D, Input=RGB and TVL1 flow2021.05 | 41 | — | — | 24.4 | — | 33.5 | — | 26.9 | 19.8 | — | 10.8 | |
| DGAMSupervision=Weakly2021.07 | 41 | 23.5 | 5.3 | 24.4 | — | — | — | — | — | — | — | |
| HAM-NetSupervision=Weakly2021.07 | 41 | 24.8 | 5.3 | 25.1 | — | — | — | — | — | — | — | |
| DGAM2023.05 | 41 | 23.5 | 5.3 | 24.4 | — | — | — | — | — | — | — | |
| ASLSupervision=Weakly Supervised, Backbone=I3D, Input=RGB and TVL1 flow2021.05 | 40.2 | — | — | 25.8 | — | 34.6 | — | 29.4 | 22.5 | — | 12.1 | |
| ACSNetSupervision=Weak2021.03 | 40.1 | 26.1 | 6.8 | 26 | — | — | — | — | — | — | — | |
| ACSNetSupervision=Weakly2021.07 | 40.1 | 26.1 | 6.8 | 26 | — | — | — | — | — | — | — | |
| Gong et al. (2020)Supervision Level=Weak, Action Counts=False2020.06 | 40 | 25 | 4.6 | 24.6 | — | — | — | — | — | — | — | |
| TCAM2021.04 | 40 | 25 | 4.6 | 24.6 | — | — | — | — | — | — | — | |
| Gong et al.Supervision=Weakly2021.07 | 40 | 25 | 4.6 | 24.6 | — | — | — | — | — | — | — | |
| TCAM2023.05 | 40 | 25 | 4.6 | 24.6 | — | — | — | — | — | — | — | |
| ActionBytesSupervision=Weakly2021.07 | 39.4 | — | — | — | — | — | — | — | — | — | — | |
| RefineLocSupervision=Weakly2021.07 | 38.7 | 22.6 | 5.5 | 23.2 | — | — | — | — | — | — | — | |
| BaS-NetSupervision=Weak2019.11 | 38.5 | 24.2 | 5.6 | 24.3 | — | — | — | — | — | — | — | |
| BaS-Net (2020)Supervision Level=Weak, Action Counts=False2020.06 | 38.5 | 24.2 | 5.6 | 24.3 | — | — | — | — | — | — | — | |
| BaSNetSupervision=Weak2021.03 | 38.5 | 24.2 | 5.6 | 24.3 | — | — | — | — | — | — | — | |
| Bas-Net2021.04 | 38.5 | 24.2 | 5.6 | 24.3 | — | — | — | — | — | — | — | |
| BaSNetSupervision=Weakly Supervised, Backbone=I3D, Input=RGB and TVL1 flow2021.05 | 38.5 | — | — | 24.3 | — | — | — | — | — | — | — | |
| BaSNetSupervision=Weakly2021.07 | 38.5 | 24.2 | 5.6 | 24.3 | — | — | — | — | — | — | — | |
| BaS-Net2023.05 | 38.5 | 24.2 | 5.6 | 24.3 | — | — | — | — | — | — | — | |
| SF-NetSupervision=Single-frame2020.03 | 37.8 | — | — | 22.8 | — | — | — | 24.6 | — | — | 10.3 | |
| RPN (2020)Supervision Level=Weak, Action Counts=False2020.06 | 37.6 | 23.9 | 5.4 | 23.3 | — | — | — | — | — | — | — | |
| TSCN (2020)Supervision Level=Weak, Action Counts=False2020.06 | 37.6 | 23.7 | 5.7 | 23.6 | — | — | — | — | — | — | — | |
| Huang et al.2021.04 | 37.6 | 23.9 | 5.4 | 23.3 | — | — | — | — | — | — | — | |
| TSCN2021.04 | 37.6 | 23.7 | 5.7 | 23.6 | — | — | — | — | — | — | — | |
| TSCNSupervision=Weakly Supervised, Backbone=I3D, Input=RGB and TVL1 flow2021.05 | 37.6 | — | — | 23.6 | — | — | — | 23.1 | — | — | — | |
| TSCNSupervision=Weakly2021.07 | 37.6 | 23.7 | 5.7 | 23.6 | — | — | — | — | — | — | — | |
| TSCN2023.05 | 37.6 | 23.7 | 5.7 | 23.6 | — | — | — | — | — | — | — | |
| TSCNFeatures=I3D2020.10 | 37.6 | 23.7 | 5.7 | 23.6 | — | — | — | — | — | — | — | |
| EM-MIL (2020)Supervision Level=Weak, Action Counts=False2020.06 | 37.4 | — | — | 20.3 | — | — | — | — | — | — | — | |
| EM-MILSupervision=Weakly-Supervised2020.03 | 37.4 | — | — | 20.3 | — | — | — | 23.1 | — | — | 2 | |
| EM-MIL2021.04 | 37.4 | 23.1 | 2 | 20.3 | — | — | — | — | — | — | — | |
| EM-MILSupervision=Weakly Supervised, Backbone=I3D, Input=RGB and TVL1 flow2021.05 | 37.4 | — | — | 20.2 | — | — | — | 23.1 | — | — | 2 | |
| EM-MIL2023.05 | 37.4 | 23.1 | 2 | 20.3 | — | — | — | — | — | — | — | |
| 3C-NetSupervision=Weak2020.03 | 37.2 | — | — | 21.7 | — | — | — | 23.7 | — | — | 9.2 | |
| 3C-Net (2019)Supervision Level=Weak, Action Counts=True2020.06 | 37.2 | — | — | 21.7 | — | — | — | — | — | — | — | |
| 3C-NetSupervision=Weakly-Supervised2020.03 | 37.2 | — | — | 21.7 | — | — | — | 23.7 | — | — | 9.2 | |
| 3C-Net2021.04 | 37.2 | 23.7 | — | 21.7 | — | — | — | — | — | — | — | |
| CleanNetSupervision=Weak2020.03 | 37.1 | 20.3 | 5 | 21.6 | 33.4 | 29.9 | 26.7 | 23.4 | 17.2 | 13.9 | 9.2 | |
| CleanNet (2019)Supervision Level=Weak, Action Counts=False2020.06 | 37.1 | 20.3 | 5 | 21.6 | — | — | — | — | — | — | — | |
| Liu et al.Supervision=Weakly-Supervised2020.03 | 37.1 | — | — | 21.6 | — | — | — | 23.4 | — | — | 9.2 | |
| CleanNetSupervision=Weak2021.03 | 37.1 | 20.3 | 5 | 21.6 | — | — | — | — | — | — | — | |
| Clean-Net2021.04 | 37.1 | 20.3 | 5 | 21.6 | — | — | — | — | — | — | — | |
| CleanNetSupervision=Weakly Supervised, Backbone=I3D, Input=RGB and TVL1 flow2021.05 | 37.1 | — | — | 21.6 | — | 29.9 | — | 23.4 | 17.2 | — | 9.2 | |
| W-TALCSupervision=Weak2019.11 | 37 | — | — | 18 | — | — | — | — | — | — | — | |
| W-TALCSupervision=Weak2020.03 | 37 | 12.7 | 1.5 | 18 | 33.5 | 30.4 | 25.7 | 14.6 | 10 | 7 | 4.2 | |
| W-TALCSupervision=Weak2020.03 | 37 | — | — | 18 | — | — | — | 14.6 | — | — | — | |
| W-TALC (2018)Supervision Level=Weak, Action Counts=False2020.06 | 37 | 12.7 | 1.5 | 18 | — | — | — | — | — | — | — | |
| W-TALCSupervision=Weakly-Supervised2020.03 | 37 | — | — | 18 | — | — | — | 14.6 | — | — | 4.2 | |
| W-TALCSupervision=Weak2021.03 | 37 | 12.7 | 1.5 | 18 | — | — | — | — | — | — | — | |
| WTALC2021.04 | 37 | 14.6 | — | 18 | — | — | — | — | — | — | — | |
| WTALC2023.05 | 37 | 14.6 | — | 18 | — | — | — | — | — | — | — | |
| W-TALCFeatures=I3D2020.10 | 37 | — | — | 18 | — | — | — | — | — | — | — | |
| Liu et al.Supervision=Weak2019.11 | 36.8 | 22 | 5.6 | 22.4 | — | — | — | — | — | — | — | |
| Liu et al.Supervision=Weak2020.03 | 36.8 | 22 | 5.6 | 22.4 | — | — | — | — | — | — | — | |
| Liu et al.Supervision=Weak2020.03 | 36.8 | — | — | 22.4 | — | — | — | — | — | — | — | |
| Liu et al. (2019)Supervision Level=Weak, Action Counts=False2020.06 | 36.8 | 22.9 | 5.6 | 22.4 | — | — | — | — | — | — | — | |
| Liu et al.Supervision=Weak2021.03 | 36.8 | 22 | 5.6 | 22.4 | — | — | — | — | — | — | — | |
| CMCS2021.04 | 36.8 | 22 | 5.6 | 22.4 | — | — | — | — | — | — | — | |
| CMCSSupervision=Weakly Supervised, Backbone=I3D, Input=RGB and TVL1 flow2021.05 | 36.8 | — | — | — | — | — | — | — | — | — | — | |
| CMCSSupervision=Weakly†2021.07 | 36.8 | 22 | 5.6 | 22.4 | — | — | — | — | — | — | — | |
| CMCS2023.05 | 36.8 | 22 | 5.6 | 22.4 | — | — | — | — | — | — | — | |
| Liu et al.Features=I3D2020.10 | 36.8 | 22 | 5.6 | 22.4 | — | — | — | — | — | — | — | |
| 3C-NetSupervision=Weak2020.03 | 35.4 | — | — | 21.1 | — | — | — | 22.9 | — | — | 8.5 | |
| 3CNetSupervision=Weakly Supervised, Backbone=I3D, Input=RGB and TVL1 flow2021.05 | 35.4 | — | — | 21.1 | — | — | — | 22.9 | — | — | 8.5 | |
| 3C-NetSupervision=Weakly†2021.07 | 35.4 | 22.9 | 8.5 | 21.1 | — | — | — | — | — | — | — | |
| TSMSupervision=Weak2020.03 | 28.3 | 17 | 3.5 | 17.1 | 26 | 23.6 | 21.2 | 18.9 | 14 | 11.1 | 7.5 | |
| TSM (2019)Supervision Level=Weak, Action Counts=False2020.06 | 28.3 | 17 | 3.5 | 17.1 | — | — | — | — | — | — | — | |
| TSMSupervision=Weakly-Supervised2020.03 | 28.3 | — | — | 17.1 | — | — | — | 18.9 | — | — | 7.5 | |
| TSMSupervision=Weak2021.03 | 28.3 | 17 | 3.5 | 17.1 | — | — | — | — | — | — | — | |
| TSM2021.04 | 28.3 | 17 | 3.5 | — | — | — | — | — | — | — | — | |
| TSMSupervision=Weakly Supervised, Backbone=I3D, Input=RGB and TVL1 flow2021.05 | 28.3 | — | — | 17.1 | — | 23.6 | — | 18.9 | 14 | — | 7.5 | |
| AutoLocSupervision=Weak2019.11 | 27.3 | 15.1 | 3.3 | 16 | — | — | — | — | — | — | — | |
| AutoLocSupervision=Weak, Feature Extractor=Weaker2020.03 | 27.3 | 15.1 | 3.3 | 16 | 24.9 | 22.5 | 19.9 | 17.5 | 13 | 10 | 6.8 | |
| AutoLocSupervision=Weak2020.03 | 27.3 | — | — | 16 | — | — | — | 17.5 | — | — | 6.8 | |
| AutoLoc (2018)Supervision Level=Weak, Action Counts=False2020.06 | 27.3 | 15.1 | 3.3 | 16 | — | — | — | — | — | — | — | |
| AutolocSupervision=Weakly-Supervised2020.03 | 27.3 | — | — | 16 | — | — | — | 17.5 | — | — | 6.8 |