DISTORT-AND-RECOVER-CVPR18

Code for "Distort-and-Recover: Color Enhancement with Deep Reinforcement Learning", CVPR18

Overview

You can pull&use the docker image from pjc0309/default_setup:latest to run the code. (TensorFlow 0.11.0rc, and other packages such as numpy/scipy)
Before training,prepare MIT5K train/test images in separate folders (train/raw/, train/target/, test/raw/, test/target/). And edit the path in main.py accordingly.
run_train.sh starts training.
Use parse_test.py to parse the test results. (edit the paths accordingly)
The training speed (iterations per second) should be between 20~40 it/sec. (When trained on i5-6600 and GTX 1080)

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
checkpoints		checkpoints
tools		tools
.gitignore		.gitignore
README.md		README.md
action.py		action.py
action_set.py		action_set.py
convert.py		convert.py
copy_target.py		copy_target.py
crawl_shutterstock.py		crawl_shutterstock.py
get_global_feature.py		get_global_feature.py
get_hist.py		get_hist.py
main.py		main.py
model_vgg.py		model_vgg.py
parse_test.py		parse_test.py
q_network.py		q_network.py
run_test.sh		run_test.sh
run_train.sh		run_train.sh
utils.py		utils.py