/visual_attention_consistency

Visual Attention Consistency Under Image Transforms for Multi-Label Image Classification

Primary LanguagePython

Visual Attention Consistency

This repository is for the following paper:

@InProceedings{Guo_2019_CVPR,
author = {Guo, Hao and Zheng, Kang and Fan, Xiaochuan and Yu, Hongkai and Wang, Song},
title = {Visual Attention Consistency Under Image Transforms for Multi-Label Image Classification},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2019}
}

Note:

This is just a preliminary version for early access. I will complete it as soon as I can.

WIDER Attribute Dataset

To run this code, you need to specify the WIDER_DATA_DIR (WIDER dataset "Image"), WIDER_ANNO_DIR (WIDER dataset annotations) in "configs.py"and the argument of model_dir (path to save checkpoints). Then, run the command (with PyTorch installed): python main.py.

Note that you may need to use the specific PyTorch version: 0.3.1.

PA-100K Dataset

(To be integrated)

MS-COCO

(To be integrated)

For those who want to test the proposed method on MS-COCO dataset, the source codes which I used for experiments are temporarily uploaded in the ./tmp/ folder. I will do code cleanup later.

Select the checkpoint that produces the best mAP.

You can also evaluate the predictions with code at: https://github.com/zhufengx/SRN_multilabel, which would produce a slightly better (~0.1%) performance.

Datasets

  1. WIDER Attribute Dataset: http://mmlab.ie.cuhk.edu.hk/projects/WIDERAttribute.html
  2. PA-100K: (will be supported later)
  3. MS-COCO: (will be supported later)