https://github.com/bcmi/Composite-Image-Evaluation

https://github.com/bcmi/Composite-Image-Evaluation

image-composition

Last synced: 2 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/bcmi/Composite-Image-Evaluation
• Owner: bcmi
• Created: 2023-08-15T13:08:57.000Z (11 months ago)
• Default Branch: main
• Last Pushed: 2023-11-10T13:48:14.000Z (8 months ago)
• Last Synced: 2024-01-27T20:46:25.530Z (5 months ago)
• Topics: image-composition
• Homepage:
• Size: 4.88 KB
• Stars: 5
• Watchers: 4
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Lists

• Awesome-Image-Composition - Composite-Image-Evaluation
• Awesome-Generative-Image-Composition - Composite-Image-Evaluation
• Awesome-Generative-Image-Composition?tab=readme-ov-file - Composite-Image-Evaluation

README

        
# Composite-Image-Evaluation

Here are some possible evaluation metrics to evaluate the quality of composite images from different aspects. 

+ Evaluate whether the foreground is harmonious with background.

  + Harmony score: use [illumination encoder](https://github.com/bcmi/BargainNet-Image-Harmonization) to extract the illumination codes from foreground and background, and measure their similarity.

 

  + Inharmony hit: use [inharmonious region localization model](https://github.com/bcmi/MadisNet-Inharmonious-Region-Localization) to detect the inharmonious region, and calculate the overlap (e.g., IoU) between detected region and foreground region. 

+ Evaluate whether the foreground object placement is reasonable.

  + OPA score: use [object placement assessment model](https://github.com/bcmi/Object-Placement-Assessment-Dataset-OPA) to predict the accuracy of object placement. 

+ Evaluate whether the foreground is compatible with background in terms of geometry and semantics.

  + FOS score: use [foreground object search model](https://github.com/bcmi/Foreground-Object-Search-Dataset-FOSD) to calculate the compatibility score between foreground and background in terms of geometry and semantics. 

+ Evaluate the fidelity of foreground, i.e., whether the synthesized foreground is similar to the input foreground.

  + Clip score: use [CLIP](https://github.com/openai/CLIP) to extract the embeddings from the input foreground image and the generated foreground patch, and measure their similarity.

 

  + Dino score: use [DINO](https://github.com/facebookresearch/dino) to measure the average cosine similarity between the input and generated foreground.  

+ Evaluate the over quality of foreground or the whole composite image.

  + FID: use pretrained image encoder (*e.g.*, InceptionNet, CLIP) to extract the embeddings from real images and generated images, and measure their [Fréchet Inception Distance](https://github.com/mseitzer/pytorch-fid).

  + QS: use [quality score](https://github.com/cientgu/GIQA) to measure the quality of each single generated image, and compute average score.