Deep Learning Temporal Action Detection
Last updated: 2019/9/25
Performance Table
| Detector | THUMOS (mAP@IoU=0.5) | ANET (mAP@IoU=0.5) | Speed | Published In |
|---|---|---|---|---|
| LAF | 4.4 | - | ACMMM'15 | |
| RNN & RL | 17.1 | - | CVPR'16 | |
| PSDF | 18.8 | - | CVPR'16 | |
| SSAD | 24.6 | - | CVPR'16 | |
| SCNN | 19.0 | - | CVPR'16 | |
| SCNNv2 | 19.0 | - | WACV'16 | |
| CBR | 31.0 | - | BMVC'17 | |
| SMS | 14.8 | - | CVPR'17 | |
| UntrimmedNets | 13.7 | 7.2 | CVPR'17 | |
| TCN | 25.6 | 23.6 | ICCV'17 | |
| ETE SSTAD | 29.2 | - | BMVC'17 | |
| CDC | 23.3 | 22.9 | 500 | ICCV'17 |
| SMS | 17.8 | - | CVPR'17 | |
| SCC | 19.3 | - | CVPR'17 | |
| TAG | 28.3 | - | CVPR'17 | |
| SST | 19.3 | - | CVPR'17 | |
| SSTAD | 24.6 | - | ACMMM'17 | |
| TURN-TAP | 25.6 | - | 880 | ICCV'17 |
| R-C3D | 28.9 | 16.7 | ICCV'17 | |
| TAG+SSN | 29.1 | 28.3 | ICCV'17 | |
| ETP | 34.2 | - | CVPR'18 | |
| TAL-Net | 42.0 | 38.2 | CVPR'18 | |
| STPN | 16.9 | 29.3 | CVPR'18 | |
| TPN | 27.6 | - | AAAI'18 | |
| SAP | 27.7 | - | AAAI'18 | |
| WSTAD | 15.9 | 27.3 | ACMMM'18 | |
| BSN | 36.9 | - | ECCV'18 | |
| AutoLoc | 21.2 | 27.3 | ECCV'18 | |
| W-TALC | 22.8 | 37.0 | ECCV'18 | |
| CPMN | 16.1 | 39.3 | ACMMM'18 | |
| STAR | 23.0 | 31.1 | AAAI'19 | |
| BMN | 38.8 | 39.4 | ICCV'19 | |
| MGG | 39.9 | - | ICCV'19 |
Dataset
-
[THUMOS] THUMOS Challenge 2014 |
[Homepage] -
[Activity-Net] A Large-Scale Video Benchmark for Human Activity Understanding |
[Homepage] -
[COIN] A Large-scale Dataset for Comprehensive Instructional Video Analysis |
[Homepage]
2015
- [LAF] Temporal Localization of Fine-Grained Actions in Videos by Domain Transfer from Web Images | [ACMMM'15] |
[pdf]
2016
-
[RNN & RL] End-to-end learning of action detection from frame glimpses in videos | [CVPR'16] |
[pdf] -
[PSDF] Temporal Action Localization with Pyramid of Score Distribution Features | [CVPR'16] |
[pdf] -
[SCNN] Temporal action localization in untrimmed videos via multi-stage CNNs | [CVPR'16] |
[pdf] -
[SCNNv2] Efficient Action Detection in Untrimmed Videos via Multi-Task Learning | [WACV'16] |
[pdf]
2017
-
[SSAD] Single Shot Temporal Action Detection | [ACMMM'17] |
[pdf] -
[CBR] Cascaded Boundary Regression for Temporal Action Detection | [BMVC'17] |
[pdf] -
[SMS] Temporal Action Localization by Structured Maximal Sums | [CVPR'17] |
[pdf] -
[UntrimmedNets] UntrimmedNets for Weakly Supervised Action Recognition and Detection | [CVPR'17] |
[pdf] -
[TCN] Temporal Context Network for Activity Localization in Videos | [ICCV'17] |
[pdf] -
[SCC] Semantic Context Cascade for Efficient Action Detection | [CVPR'17] |
[pdf] -
[SSTAD] Single Shot Temporal Action Detection | [ACMMM'17] |
[pdf] -
[ETE SSTAD] End-to-End, Single-Stream Temporal Action Detection in Untrimmed Videos | [BMVC'17] |
[pdf] -
[SST] Single-stream temporal action proposals | [CVPR'17] |
[pdf] -
[TURN-TAP] Temporal Unit Regression Network for Temporal Action Proposals | [ICCV'17] |
[pdf] -
[Hide-and-Seek] Hide-and-Seek: Forcing a Network to be Meticulous for Weakly-supervised Object and Action Localization | [ICCV'17] |
[pdf] -
[TAG] A Pursuit of Temporal Accuracy in General Activity Detection | [CVPR'17] |
[pdf]
2018
-
[ETP] Precise Temporal Action Localization by Evolving Temporal Proposals | [CVPR'18] |
[pdf] -
[TAL-Net] Rethinking the Faster R-CNN Architecture for Temporal Action Localization | [CVPR'18] |
[pdf] -
[STPN] Weakly Supervised Action Localization by Sparse Temporal Pooling Network| [CVPR'18] |
[pdf] -
[TPN] Exploring Temporal Preservation Networks for Precise Temporal Action Localization | [AAAI'18] |
[pdf] -
[SAP] A Self-Adaptive Proposal Model for Temporal Action Detection based on Reinforcement Learning | [AAAI'18] |
[pdf] -
[WSTAD] Step-by-step Erasion, One-by-one Collection: A Weakly Supervised Temporal Action Detector | [ACMMM'18] |
[pdf] -
[BSN] Boundary Sensitive Network for Temporal Action Proposal Generation | [ECCV'18] |
[pdf] -
[AutoLoc] AutoLoc: Weakly-supervised Temporal Action Localization in Untrimmed Videos | [ECCV'18] |
[pdf] -
[W-TALC] W-TALC: Weakly-supervised Temporal Activity Localization and Classification | [ECCV'18] |
[pdf] -
[CPMN] Cascaded Pyramid Mining Network for Weakly Supervised Temporal Action Localization | [ACCV'18] |
[pdf]