Home

Awesome

GaitGL

NOTE

This repo is based on GaitSet

Prerequisites

Dataset & Preparation

Download OU-MVLP Dataset.

!!! ATTENTION !!! ATTENTION !!! ATTENTION !!!

Before training or test, please make sure you have prepared the dataset by this two steps:

Pretreatment

pretreatment_oumvlp.py uses the alignment method in this paper. Pretreatment your dataset by

python pretreatment_oumvlp.py --input_path='root_path_of_raw_dataset' --output_path='root_path_for_output'

Train

Train a model by

python train.py

'batch_size': (32, 8), 'frame_num': 30, 'total_iter': 250000.The learning rate is 1e − 4 in the first 150K iterations, and then is changed into 1e − 5 for the rest of 100K iterations.

Evaluation

Evaluate the trained model by

python test_oumvlp.py

CAISA-E

Dataset & Preparation

Function generate_test_gallery() generate_train_gallery() generate_test_probe() from pt_casiae.py

Train

OUMVLP Pre-training parameters need to be added. Train a model by

python train.py

'batch_size': (12, 8), 'frame_num': 64, 'total_iter': 15000. The learning rate is 1e − 4 in the first 10K iterations, and then is changed into 1e − 5 for the rest of 5K iterations.

Test

Training parameters. Test a model by using Function testout() from pt_casiae.py

python pt_casiae.py

Citation

Please cite these papers in your publications if it helps your research:

@article{linlearning,
  title={Learning Effective Representations from Global and Local Features for Cross-View Gait Recognition},
  author={Lin, Beibei and Zhang, Shunli and Yu, Xin and Kong, Chuihan and Wan, Chenwei}
}