Skip to content
/ GCPNet Public

Scene Parsing with Global Context Embedding, ICCV 2017

Notifications You must be signed in to change notification settings

hfslyc/GCPNet

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Scene Parsing with Global Context Embedding

This repo is the caffe implementation of the following paper:

Scene Parsing with Global Context Embedding
Wei-Chih Hung, Yi-Hsuan Tsai, Xiaohui Shen, Zhe Lin, Kalyan Sunkavalli, Xin Lu, and Ming-Hsuan Yang. In ICCV, 2017.

Please cite our paper if you find it useful for your research.

@inproceedings{hung2017scene,
  title={Scene Parsing With Global Context Embedding},
  author={Hung, Wei-Chih and Tsai, Yi-Hsuan and Shen, Xiaohui and Lin, Zhe and Sunkavalli, Kalyan and Lu, Xin and Yang, Ming-Hsuan},
  booktitle={IEEE International Conference on Computer Vision (ICCV)},
  year={2017}
}

Prerequisite

  • DeepLab-v2 caffe. You will need the this updated version [link] for most recent machine setups.
  • A GPU with at least 12GB

Test on ADE20k validation set

  • Download the ADE20k dataset and put it in data/.

The directories should be like this:

./data/ADE20k/annotations/validation
             /images/validation
  • Download pretrained model
bash get_model.sh
  • Download precomputed context features/priors of ADE20k val set.
bash get_prior.sh
  • Execute evaluation script:
python eval.py --prototxt prototxt/ade20k_val.prototxt \
               --model models/ade20k_full.caffemodel \
               --save-dir results/ade20k/val/ \
               --gpu 0

The result images will be saved at results/ade20k/val/.

About

Scene Parsing with Global Context Embedding, ICCV 2017

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published