【资源】时序行为检测相关资源列表
source link: https://bbs.cvmart.net/articles/526
Go to the source link to view the article. You can view the picture content, updated content and better typesetting reading experience. If the link is broken, please click the button below to view the snapshot at that time.
作者:Rheelt
项目:Materials-Temporal-Action-Detection
本文汇总了时序行为检测相关的资源列表
更多Awsome Github资源请关注:【Awsome】GitHub 资源汇总
Papers: temporal action proposals & detection
- (TGM) Temporal Gaussian Mixture Layer for Videos (ICML 2019) CODE.pytorch
- (MGG) Multi-granularity Generator for Temporal Action Proposal (CVPR 2019)
- (GTAN) Gaussian Temporal Awareness Networks for Action Localization (CVPR 2019)
Papers: weakly temporal action detection
- (CMCS) Completeness Modeling and Context Separation for Weakly Supervised Temporal Action Localization (CVPR19)CODE.pytorch
- Weakly-Supervised Temporal Localization via Occurrence Count Learning (ICML 2019)
- (MAAN) Marginalized Average Attentional Network for Weakly-Supervised Learning (ICLR2019)CODE.pytorch
- (WSGN) Weakly Supervised Gaussian Networks for Action Detection (Arxiv 2019.4)
- (RefineLoc) RefineLoc: Iterative Refinement for Weakly-Supervised Action Localization (Arxiv 2019.4)
- (STAR) Segregated Temporal Assembly Recurrent Networks for Weakly Supervised Multiple Action Detection (AAAI 2019)
- (TSRNet) Learning Transferable Self-attentive Representations for Action Recognition in Untrimmed Videos with Weak Supervision (AAAI 2019)
- (StepByStep) Step-by-step Erasion, One-by-one Collection: AWeakly Supervised Temporal Action Detector (MM 2018)
- (W-TALC) W-TALC: Weakly-supervised Temporal Activity Localization and Classification (ECCV 2018) CODE.pytorch
- (AutoLoc) AutoLoc: Weakly-supervised Temporal Action Localization in Untrimmed Videos (ECCV 2018) CODE.caffe
- (STPN) Weakly Supervised Action Localization by Sparse Temporal Pooling Network (CVPR 2018) CODE.tensorflow.unofficial
- (H&S) Hide-and-seek: Forcing a network to be meticulous for weakly-supervised object and action localization (ICCV 2017)
- (UNet) UntrimmedNets for Weakly Supervised Action Recognition and Detection (CVPR 2017) CODE.caffe
Features: Download link
THUMOS14
- C3D: link
- I3D: Video is sampled at 25 frames per second. 16 frames as a video unit. link
- UNet: link
- ANet2016-cuhk(4096dims): 6 frames as a video unit. link
- ANet2016-cuhk(3072dims): 5 frames as a video unit. link
ActivityNet v1.2
ActivityNet v1.3
- C3D: link
- ANet2016-cuhk(400dims): 16 frames as a video unit. link
- I3D: 16 frames as a video unit. link
- ANet2016-cuhk(3072dims): 16 frames as a video unit. link
Benchmark Results (THUMOS14 Results)
These methods are listed in chronological order.
Method Feature IoU-> 0.1 0.2 0.3 0.4 0.5 0.6 0.7 MAAN I3D
59.8 50.8 41.1 30.6 20.3 12.0 6.9 CMCS I3D
57.4 50.8 41.2 32.1 23.1 15.0 7.0 WSGN I3D
55.3 47.6 38.9 30.0 21.1 13.9 8.3 RefineLoc UNet
33.9
22.1
6.1 STAR I3D
68.8 60.0 48.7 34.7 23.0
TSRNet 2-Stream(ResNet101)
55.9 46.9 38.3 28.1 18.6 11.0 5.59 StepByStep TSN
45.8 39.0 31.1 22.5 15.9
W-TALC UNet
49.0 42.8 32.0 26.0 18.8
6.2 W-TALC I3D
55.2 49.6 40.1 31.1 22.8
7.6 AutoLoc UNet
35.8 29.0 21.2 13.4 5.8 STPN I3D
52.0 44.7 35.5 25.8 16.9 9.9 4.3 H&S C3D
36.4 27.8 19.5 12.7 6.8
UNet UNet
44.4 37.7 28.2 21.1 13.7
推荐阅读:
【资源】常用的语义分割架构结构综述以及代码复现
【资源】视频分析 / 多模态学习论文、代码、数据集大列表
微信公众号: 极市平台(ID: extrememart )
每天推送最新CV干货
Recommend
About Joyk
Aggregate valuable and interesting links.
Joyk means Joy of geeK