hichoe95/Artifact-Detection-and-Sequential-Ablation

[IJCAI-2022] Can We Find Neurons that Cause Unrealistic Images in Deep Generative Networks?

22
/ 100
Experimental

This tool helps researchers and practitioners working with deep generative models like GANs to identify and fix "artifacts" – unrealistic or defective elements – in the images these models create. It takes a pre-trained generative model and dataset, then outputs information about which internal components (neurons) are causing these artifacts and can generate corrected, higher-quality images. It's for anyone who trains or uses generative AI models and needs to ensure their outputs are photorealistic.

No commits in the last 6 months.

Use this if you are generating images with GANs and notice recurring imperfections or unrealistic details you want to automatically detect and correct.

Not ideal if you are working with generative models other than PGGAN or StyleGAN2, or if your primary goal is not image generation.

generative-ai image-synthesis model-debugging computer-vision AI-safety
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 6 / 25
Maturity 16 / 25
Community 0 / 25

How are scores calculated?

Stars

24

Forks

Language

Jupyter Notebook

License

MIT

Last pushed

Nov 19, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/generative-ai/hichoe95/Artifact-Detection-and-Sequential-Ablation"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.