Graph Contrastive Learning for Skeleton-based Action Recognition
Authors: Xiaohu Huang, Hao Zhou, Jian Wang, Haocheng Feng, Junyu Han, Errui Ding, Jingdong Wang, Xinggang Wang, Wenyu Liu, Bin Feng
ICLR 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We combine Skeleton GCL with three GCNs (2S-ACGN, CTR-GCN, and Info GCN), and achieve consistent improvements on NTU60, NTU120, and NW-UCLA benchmarks.4 EXPERIMENTS |
| Researcher Affiliation | Collaboration | Xiaohu Huang 1, 2 Hao Zhou 2 Jian Wang 2 Haocheng Feng 2 Junyu Han 2 Errui Ding 2 Jingdong Wang 2 Xinggang Wang 1 Wenyu Liu 1 Bin Feng 1 1 School of EIC, Huazhong University of Science & Technology 2 Department of Computer Vision Technology (VIS), Baidu Inc., China |
| Pseudocode | No | The paper does not contain any clearly labeled 'Pseudocode' or 'Algorithm' block. |
| Open Source Code | Yes | The source code will be available at https://github.com/Oliver Hxh/Skeleton GCL. |
| Open Datasets | Yes | NTU RGB+D (NTU60) (Shahroudy et al., 2016) is a large-scale skeleton-based action recognition dataset... NTU RGB+D 120 (NTU120) (Liu et al., 2019) is an extension of NTU RGB+D dataset... Northwestern-UCLA (NW-UCLA) dataset (Wang et al., 2014) contains 1494 sequences from 10 action classes. |
| Dataset Splits | Yes | Generally, two protocols are used to evaluate the performances: (1) cross-subject (X-Sub): train data are performed by 20 subjects, and test data are performed by other 20 subjects. (2) cross-view (X-View): train data from cameras 2 and 3, and test data from camera 1. |
| Hardware Specification | Yes | All experiments are conducted using a single NVIDIA V100 GPU. |
| Software Dependencies | No | The paper does not provide specific version numbers for software dependencies or libraries used in the experiments. |
| Experiment Setup | Yes | P, the number of stored instances for each class in MIns, is set as 684 on NTU60 and NTU120, and 342 on NW-UCLA. The dimension of graph vector Cg is set to 256. For all datasets, the number of sampling examples K+ H , K H , and K R are set as 128, 512, and 512, respectively. For different models used in different modalities, we experiment with temperature τ of 0.5, 0.8, 1.0, and 1.5, and choose the best one. The hyper-parameter α for momentum updating is set as 0.85. |