Brain-Cog-Lab/AbuttingGratingIllusion

The code for "Challenging Deep Learning Models with Image Distortion based on the Abutting Grating Illusion"

34
/ 100
Emerging

This project helps evaluate how robust your deep learning models are at recognizing objects when images are subtly distorted by visual illusions, specifically the 'abutting grating illusion'. It takes existing image datasets and applies this unique distortion, then measures how well various deep learning models perform compared to human perception. Researchers in AI safety, cognitive science, or model interpretability would use this to understand model limitations.

No commits in the last 6 months.

Use this if you need to test the robustness of your deep learning vision models against human-like perceptual distortions, rather than just mathematical noise or transformations.

Not ideal if you're looking for general image augmentation techniques or solutions for standard image classification problems without a focus on cognitive robustness.

AI Safety Cognitive AI Model Robustness Perceptual Science Deep Learning Evaluation
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 5 / 25
Maturity 16 / 25
Community 13 / 25

How are scores calculated?

Stars

9

Forks

2

Language

Jupyter Notebook

License

MIT

Last pushed

Mar 01, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/Brain-Cog-Lab/AbuttingGratingIllusion"

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