SiT-MLP

This is the official implementation of our paper SiT-MLP: A Simple MLP with Point-wise Topology Feature Learning for Skeleton-based Action Recognition

Note: Our approch is MLP-based and GCN-free. The graph folder is adopted for different modality.

Abstract

Graph convolution networks (GCNs) have achieved remarkable performance in skeleton-based action recognition. However, previous GCN-based methods rely on elaborate human priors excessively and construct complex feature aggregation mechanisms, which limits the generalizability and effectiveness of networks. To solve these problems, we propose a novel Spatial Topology Gating Unit (STGU), an MLP-based variant without extra priors, to capture the co-occurrence topology features that encode the spatial dependency across all joints. In STGU, to learn the point-wise topology features, a new gate-based feature interaction mechanism is introduced to activate the features point-to-point by the attention map generated from the input sample. Based on the STGU, we propose the first MLP-based model, SiT-MLP, for skeleton-based action recognition in this work. Compared with previous methods on three large-scale datasets, SiT-MLP achieves competitive performance. In addition, SiT-MLP reduces the parameters significantly with favorable results.

Experiment

Model	Parameters	FLOPs	Accuracy
2S-AGCN	3.5M	3.9G	95.1
CTR-GCN	1.4M	1.8G	96.8
InfoGCN	1.5M	1.7G	96.7
SiT-MLP	0.6M	0.7G	96.8

Comparsion of performance and parameter size on X-sub benchmark of NTU RGB+D 60 dataset. We report the accuracy as performance on the vertical axis. The closer to the top-left, the better.

SiT-MLP architecture

Prerequisites

You can install all dependencies by running pip install -r requirements.txt
Then, you need to install torchlight by running pip install -e torchlight

Data Preparation

Download datasets.

There are 3 datasets to download:

• NTU RGB+D 60 Skeleton
• NTU RGB+D 120 Skeleton
• NW-UCLA

NTU RGB+D 60 and 120

1. Request dataset here: https://rose1.ntu.edu.sg/dataset/actionRecognition
2. Download the skeleton-only datasets:
   1. nturgbd_skeletons_s001_to_s017.zip (NTU RGB+D 60)
   2. nturgbd_skeletons_s018_to_s032.zip (NTU RGB+D 120)
   3. Extract above files to ./data/nturgbd_raw

NW-UCLA

1. Download dataset from here
2. Move all_sqe to ./data/NW-UCLA

Data Processing

Directory Structure

Put downloaded data into the following directory structure:

- data/
  - NW-UCLA/
    - all_sqe
      ... # raw data of NW-UCLA
  - ntu/
  - ntu120/
  - nturgbd_raw/
    - nturgb+d_skeletons/     # from `nturgbd_skeletons_s001_to_s017.zip`
      ...
    - nturgb+d_skeletons120/  # from `nturgbd_skeletons_s018_to_s032.zip`
      ...

Generating Data

• Generate NTU RGB+D 60 or NTU RGB+D 120 dataset:

 cd ./data/ntu # or cd ./data/ntu120
 # Get skeleton of each performer
 python get_raw_skes_data.py
 # Remove the bad skeleton 
 python get_raw_denoised_data.py
 # Transform the skeleton to the center of the first frame
 python seq_transformation.py

Training & Testing

Training

• To train model on NTU60/120

# Example: training SiT-MLP on NTU RGB+D cross subject joint modality
CUDA_VISIBLE_DEVICES=0,1 python main.py --config config/nturgbd-cross-subject/mlp_joint.yaml 
# Example: training SiT-MLP on NTU RGB+D cross subject bone modality
CUDA_VISIBLE_DEVICES=0,1 python main.py --config config/nturgbd-cross-subject/mlp_bone.yaml 
# Example: training SiT-MLP on NTU RGB+D 120 cross subject joint modality
CUDA_VISIBLE_DEVICES=0,1 python main.py --config config/nturgbd120-cross-subject/mlp_joint.yaml 
# Example: training SiT-MLP on NTU RGB+D 120 cross subject bone modality
CUDA_VISIBLE_DEVICES=0,1 python main.py --config config/nturgbd120-cross-subject/mlp_bone.yaml 

• To train model on NW-UCLA

CUDA_VISIBLE_DEVICES=0,1 python main.py --config config/ucla/mlp_joint.yaml 

Testing

• To test the trained models saved in <work_dir>, run the following command:

python main.py --config <work_dir>/config.yaml --work-dir <work_dir> --phase test --save-score True --weights <work_dir>/xxx.pt

• To ensemble the results of different modalities, run

# Example: ensemble four modalities of SiT-MLP on NTU RGB+D 120 cross subject
python ensemble.py --datasets ntu120/xsub --position_ckpts work_dir/ntu120/xsub/mlp/joint work_dir/ntu120/xsub/mlp/bone --motion_ckpts work_dir/ntu120/xsub/mlp/joint_vel work_dir/ntu120/xsub/mlp/bone_vel

Acknowledgements

This repo is based on CTR-GCN. The data processing is borrowed from SGN and HCN. The code for different modality is adopted from InfoGCN.

Citation

Please cite this work if you find it useful:

@article{zhang2023sitmlp,
    author={Zhang, Shaojie and Yin, Jianqin and Dang, Yonghao and Fu, Jiajun},
    journal={IEEE Transactions on Circuits and Systems for Video Technology}, 
    title={SiT-MLP: A Simple MLP with Point-wise Topology Feature Learning for Skeleton-based Action Recognition}, 
    year={2024},
    doi={10.1109/TCSVT.2024.3386553}
}