Skip to content

Latest commit

 

History

History
129 lines (93 loc) · 4.55 KB

README.md

File metadata and controls

129 lines (93 loc) · 4.55 KB
Raw

CoFInAl: Enhancing Action Quality Assessment with Coarse-to-Fine Instruction Alignment

Kanglei Zhou   Junlin Li   Ruizhi Cai   Liyuan Wang   Xingxing Zhang   Xiaohui Liang

CoFInAl is the implementation for Action Quality Assessment (AQA) based on the paper "CoFInAl: Enhancing Action Quality Assessment with Coarse-to-Fine Instruction Alignment", which has been accepted by IJCAI 2024.

Framework

Overview

CoFInAl aims to address challenges related to domain shift and overfitting, key issues in AQA. To this end, CoFInAl strategically aligns AQA objectives with broader tasks through a coarse-to-fine classification strategy. Inspired by the two-step assessment process used by judges, CoFInAl first identifies a coarse grade and then discerns variations within each grade, enhancing interpretability within the AQA framework.

Datasets

Here are the instructions for obtaining the features and videos for the Rhythmic Gymnastics and Fis-V datasets used in our experiments:

For VST features:

  • The VST features and label files of Rhythmic Gymnastics and Fis-V datasets can be download from the GDLT repository.

For I3D features:

  • The I3D features and label files for both datasets will be released soon.

For Rhythmic Gymnastics videos:

  • Download the videos from the ACTION-NET repository.

For Fis-V videos:

  • Download the videos from the MS_LSTM repository.

Please use the above public repositories to obtain the features and videos needed to reproduce our results. Let us know if you need any clarification or have trouble accessing the data.

After downloading the Rhythmic Gymnastics dataset features and videos from the referenced repositories, preprocess the data by using rg_swinx.py.

# Choose different head to extract features like load_model or load_model_I3d
data_path = 'Path/to/Video'

orig_save = 'Path/to/Save/swintx_orig_fps25_clip{}'.format(clip_len)
pool_save = 'Path/to/Save/swintx_avg_fps25_clip{}'.format(clip_len)

# Command
python rg_swintx.py

Installation

To get started, you will need to first clone this project and then install the required dependencies.

Environments

- RTX3090 - CUDA: 11.1 - Python: 3.8+ - PyTorch: 1.10.1 + cu111

Basic packages

Install the required packages:

pip install -r requirements.txt

This will install all the required packages listed in the requirements.txt file.

Training from scratch

Using the following command to train the model:

CUDA_VISIBLE_DEVICES={device ID} python main.py \
    --video-path {path}/swintx_avg_fps25_clip32 \
    --train-label-path {path}/train.txt \
    --test-label-path {path}/test.txt  \
    --model-name ${save_name} \
    --submodel-name ${sub_model_name} \
    --action-type Ball/{action type of RG or TES/PCS} \
    --lr 1e-2 --epoch 500 \
    --n_encoder 2 --n_decoder 4 --n_query 4 --alpha 0.5 --margin 1 --lr-decay cos --decay-rate 1e-2 --dropout 0.3 \
    --loss_align 1 --activate_regular_restrictions 3

For all hyper-parameters, see scripts/train_all.sh.

Testing

Using the following command to test the model:

CUDA_VISIBLE_DEVICES={device ID} python main.py --video-path {path}/swintx_avg_fps25_clip32 \
    --train-label-path {path}/train.txt \
    --test-label-path {path}/test.txt  \
    --n_decoder 2 --n_query 4 --dropout 0.3 --test \
    --model-name ${save_name} \
    --submodel-name ${sub_model_name} \
    --action-type Ball/{action type of RG or TES/PCS} \
    --ckpt {pkl file here}

For all hyper-parameters, see scripts/test_all.sh.

Citation

Please cite this work if you find it useful:

@inproceedings{zhou2024cofinal,
  title={CoFInAl: Enhancing Action Quality Assessment with Coarse-to-Fine Instruction Alignment},
  author={Zhou, Kanglei and Li, Junlin and Cai, Ruizhi and Wang, Liyuan and Zhang, Xinxgxing and Liang Xiaohui},
  booktitle={International Joint Conference on Artificial Intelligence (IJCAI)},
  year={2024}
}

Contact

If you have any specific questions or if there's anything else you'd like assistance with regarding the code, feel free to let me know.