Source code for the following paper(arXiv link):
Learning Actor Relation Graphs for Group Activity Recognition
Jianchao Wu, Limin Wang, Li Wang, Jie Guo, Gangshan Wu
in CVPR 2019
3.x
0.4.1
data/volleyball
or data/collective
.Stage1: Fine-tune the model on single frame without using GCN.
# volleyball dataset
python scripts/train_volleyball_stage1.py
# collective dataset
python scripts/train_collective_stage1.py
Stage2: Fix weights of the feature extraction part of network, and train the network with GCN.
# volleyball dataset
python scripts/train_volleyball_stage2.py
# collective dataset
python scripts/train_collective_stage2.py
You can specify the running arguments in the python files under scripts/
directory. The meanings of arguments can be found in config.py
@inproceedings{CVPR2019_ARG,
title = {Learning Actor Relation Graphs for Group Activity Recognition},
author = {Jianchao Wu and Limin Wang and Li Wang and Jie Guo and Gangshan Wu},
booktitle = {CVPR},
year = {2019},
}