Color2Embed

Color2Embed: Fast Exemplar-Based Image Colorization using Color Embeddings Paper

This project is the simple implementation of Color2Embed. This paper maybe not be submitted to any conferences and journals and you can use it in your projects. You can help to star this repo if you think this repo is helpful.

Other recommended projects:

Temporally-Consistent-Video-Colorization

Deep Exemplar-based Colorization

Deep Video Prior

Learning Blind Video Temporal Consistency

Swapping Autoencoder

Dependencies

• Python >= 3.6 (Recommend to use Anaconda)
• PyTorch >= 1.5.0

Test

1. Clone this github repo.

git clone https://github.com/zhaohengyuan1/Color2Embed.git
cd Color2Embed

2. Pretrained models should be placed in ./experiments/ folder. VGG model also can be downloaded.

3. I have collected some test datasets used in previous papers. You can check it in the path ./test_datasets. When you use the file test_gray2color.py to test, you need to edit the input file path and pretrained weights path in this file.

python test_gray2color.py

Train

1. Prepare the training data.

cd data
sh prepare_data.sh

2. Run the train.sh. You can check train.py for more implementation details.

sh train.sh

Results

Contact

If you have any question, please email [email protected].