History

liaoxingyu e26182e6ec make lr warmup by iter Summary: change warmup way by iter not by epoch, which will make it more flexible when training small epochs		2021-01-22 11:17:21 +08:00
..
configs	make lr warmup by iter	2021-01-22 11:17:21 +08:00
fastattr	update fastreid V1.0	2021-01-18 11:36:38 +08:00
README.md	update fastreid V1.0	2021-01-18 11:36:38 +08:00
train_net.py	update fastreid V1.0	2021-01-18 11:36:38 +08:00

FastAttr in FastReID

This project provides a strong baseline for pedestrian attribute recognition.

Datasets Preparation

We use PA100k to evaluate the model's performance. You can do download dataset from HydraPlus-Net.

The training config file can be found in projects/FastAttr/config, which you can use to reproduce the results of the repo.

For example

python3 projects/FastAttr/train_net.py --config-file projects/FastAttr/configs/pa100.yml --num-gpus 4

We refer to A Strong Baseline of Pedestrian Attribute Recognition as our baseline methods and conduct the experiment with 4 GPUs. More details can be found in the config file and code.

Method	Pretrained	mA	Accu	Prec	Recall	F1
attribute baseline	ImageNet	80.50	78.84	87.24	87.12	86.78
FastAttr	ImageNet	77.57	78.03	88.39	84.98	86.65