Science Score: 54.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
✓CITATION.cff file
Found CITATION.cff file -
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
○DOI references
-
✓Academic publication links
Links to: arxiv.org -
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (14.4%) to scientific vocabulary
Repository
Basic Info
- Host: GitHub
- Owner: lizishi
- License: mit
- Language: Python
- Default Branch: main
- Size: 176 KB
Statistics
- Stars: 14
- Watchers: 4
- Forks: 0
- Open Issues: 2
- Releases: 0
Metadata Files
README.md
Efficient Action Counting with Dynamic Queries
This is the official PyTorch implementation of the paper "Efficient Action Counting with Dynamic Queries". It provides a novel perspective to tackle the Temporal Repetition Counting problem using a simple yet effective representation for action cycles, reducing the computational complexity from quadratic to linear with SOTA performance.
Installation
We build our code based on the MMaction2 project (1.3.10 version). See here for more details if you are interested. MMCV is needed before install MMaction2, which can be install with: ```shell pip install mmcv-full-f https://download.openmmlab.com/mmcv/dist/{cuversion}/{torchversion}/index.html
For example, to install the latest mmcv-full with CUDA 11.1 and PyTorch 1.9.0, use the following command:
pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu111/torch1.9.0/index.html ``` For other CUDA or pytorch version, please refer to mmcv.
Then, our code can be built by
shell
cd DeTRC
pip3 install -e .
Then, Install the 1D Grid Sampling and RoI Align operators.
shell
cd DeTRC/model
python setup.py build_ext --inplace
Data
We use the TSN feature of RepCountA and UCFRep datasets. Please refer to the guidance here.
Train
Our model can be trained with
python
python tools/train.py DeTRC/configs/repcount_tsn_feature_enc_contrastive.py --validate
We recommend to set the --validate flag to monitor the training process.
Test
If you want to test the pretrained model, please use the following code. We provide the pretrained model here.
shell
python tools/test.py DeTRC/configs/repcount_tsn_feature_enc_contrastive.py PATH_TO_CHECKPOINT
Results
| 
| 
|
| ------------------------------------------------------------ | ------------------------------------------------------------ |
Comparison with SOTA:
Citation
If you find our work useful for your project, please cite the paper as below:
@article{li2024efficient,
title={Efficient Action Counting with Dynamic Queries},
author={Li, Zishi and Ma, Xiaoxuan and Shang, Qiuyan and Zhu, Wentao and Ci, Hai and Qiao, Yu and Wang, Yizhou},
journal={arXiv preprint arXiv:2403.01543},
year={2024}
}
Owner
- Login: lizishi
- Kind: user
- Repositories: 1
- Profile: https://github.com/lizishi
Citation (CITATION.cff)
cff-version: 1.2.0 message: "If you use this software, please cite it as below." authors: - name: "MMAction2 Contributors" title: "OpenMMLab's Next Generation Video Understanding Toolbox and Benchmark" date-released: 2020-07-21 url: "https://github.com/open-mmlab/mmaction2" license: Apache-2.0
GitHub Events
Total
- Issues event: 1
- Watch event: 3
- Fork event: 1
Last Year
- Issues event: 1
- Watch event: 3
- Fork event: 1
Dependencies
- numpy *
- torch >=1.3
- docutils ==0.16.0
- einops *
- myst-parser *
- opencv-python *
- recommonmark *
- scipy *
- sphinx ==4.0.2
- sphinx_copybutton *
- sphinx_markdown_tables *
- sphinx_rtd_theme ==0.5.2
- mmcv-full >=1.3.1
- PyTurboJPEG *
- av *
- decord >=0.4.1
- einops *
- imgaug *
- librosa *
- lmdb *
- moviepy *
- onnx *
- onnxruntime *
- pims *
- timm *
- mmcv *
- titlecase *
- torch *
- torchvision *
- Pillow *
- decord *
- einops *
- matplotlib *
- numpy *
- opencv-contrib-python *
- scipy *
- coverage * test
- flake8 * test
- interrogate * test
- isort ==4.3.21 test
- pytest * test
- pytest-runner * test
- xdoctest >=0.10.0 test
- yapf * test