adaptively_weighted_attribute

This is the code for our paper: adaptively weighted multi-task deep network for person attribute classification. If you find the project useful in your research, please consider citing.

Requirements

• Caffe and pycaffe (see: Caffe installation instructions)

  Note: Caffe must be built with support for Python layers!

  # In your Makefile.config
  WITH_PYTHON_LAYER := 1

Prepare

• create model folder and data folder under current project.
• You can download the ResNet50 model and CelebA train, val, test files from google drive or baidu yun
• put the resnet_50 folder under model folder
• put the CelebA folder under data folder.

Train

• Command : python train_model.py