Skip to content

Learning Shading-aware Illumination Descriptor for Image Harmonization. Anonymous code for ECCV 2022 review.

Notifications You must be signed in to change notification settings

SID-ECCV-4751/SID

Repository files navigation

Learning Shading-aware Illumination Descriptor for Image Harmonization

Anonymous code for ECCV 2022 review

Training

We provide a mini-dataset for training our model. Specifically, our training has two stages: training the Neural Rendering Framework and training the Background Illumination Estimation Module.

Stage1: set a config file in configs/config_stage1.json, and run as following:

python train_stage1.py

Stage2: set a config file in configs/config_stage2.json, and run as following:

python train_stage2.py

Inference

Set a config file in configs/config_inference.json, and run as following:

python inference.py

Examples results:

About

Learning Shading-aware Illumination Descriptor for Image Harmonization. Anonymous code for ECCV 2022 review.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages