Skeleton Based Action Recognition on JHMDB (2D poses only)

48개 결과 · ⬇ CSV · JSON

Average accuracy of 3 splits

Rank	Model	Average accuracy of 3 splits	Accuracy	No. parameters	Paper	Code	Year
1	HT-ConvNet	86.1	86.1	1.75	Hierarchical Temporal Convolution Network:Towards Privacy-Centric Activity Recognition	Gbouna/HT-ConvNet	2024
2	DD-Net	77.2	78.0 (average of 3 split train/test)	1.82 M	Make Skeleton-based Action Recognition Model Smaller, Faster and Better	fandulu/DD-Net · BlurryLight/DD-Net-Pytorch · paty0504/SIGNTEGRATE	2019
3	JMRN (No GT pose)	68.55	–	–	Pose And Joint-Aware Action Recognition	anshulbshah/PoseAction	2020
4	PoTion	67.9	–	-	PoTion: Pose MoTion Representation for Action Recognition		2018
5	EHPI	65.5	–	–	Simple yet efficient real-time pose-based action recognition	noboevbo/ehpi_action_recognition	2019
6	Chained	56.8	–	–	Chained Multi-stream Networks Exploiting Pose, Motion, and Appearance for Action Classification and Detection	mzolfaghari/chained-multistream-networks	2017
7	HT-ConvNet	86.1	86.1	1.75	Hierarchical Temporal Convolution Network:Towards Privacy-Centric Activity Recognition	Gbouna/HT-ConvNet	2024
8	DD-Net	77.2	78.0 (average of 3 split train/test)	1.82 M	Make Skeleton-based Action Recognition Model Smaller, Faster and Better	fandulu/DD-Net · BlurryLight/DD-Net-Pytorch · paty0504/SIGNTEGRATE	2019
9	JMRN (No GT pose)	68.55	–	–	Pose And Joint-Aware Action Recognition	anshulbshah/PoseAction	2020
10	PoTion	67.9	–	-	PoTion: Pose MoTion Representation for Action Recognition		2018
11	EHPI	65.5	–	–	Simple yet efficient real-time pose-based action recognition	noboevbo/ehpi_action_recognition	2019
12	Chained	56.8	–	–	Chained Multi-stream Networks Exploiting Pose, Motion, and Appearance for Action Classification and Detection	mzolfaghari/chained-multistream-networks	2017
13	HT-ConvNet	86.1	86.1	1.75	Hierarchical Temporal Convolution Network:Towards Privacy-Centric Activity Recognition	Gbouna/HT-ConvNet	2024
14	DD-Net	77.2	78.0 (average of 3 split train/test)	1.82 M	Make Skeleton-based Action Recognition Model Smaller, Faster and Better	fandulu/DD-Net · BlurryLight/DD-Net-Pytorch · paty0504/SIGNTEGRATE	2019
15	JMRN (No GT pose)	68.55	–	–	Pose And Joint-Aware Action Recognition	anshulbshah/PoseAction	2020
16	PoTion	67.9	–	-	PoTion: Pose MoTion Representation for Action Recognition		2018
17	EHPI	65.5	–	–	Simple yet efficient real-time pose-based action recognition	noboevbo/ehpi_action_recognition	2019
18	Chained	56.8	–	–	Chained Multi-stream Networks Exploiting Pose, Motion, and Appearance for Action Classification and Detection	mzolfaghari/chained-multistream-networks	2017
19	HT-ConvNet	86.1	86.1	1.75	Hierarchical Temporal Convolution Network:Towards Privacy-Centric Activity Recognition	Gbouna/HT-ConvNet	2024
20	DD-Net	77.2	78.0 (average of 3 split train/test)	1.82 M	Make Skeleton-based Action Recognition Model Smaller, Faster and Better	fandulu/DD-Net · BlurryLight/DD-Net-Pytorch · paty0504/SIGNTEGRATE	2019