zyh-uaiaaaa/Erasing-Attention-Consistency

Official implementation of the ECCV2022 paper: Learn From All: Erasing Attention Consistency for Noisy Label Facial Expression Recognition

35
/ 100
Emerging

This project helps researchers and developers improve the accuracy of facial expression recognition models, especially when trained with imperfect or mislabeled image data. It takes in a dataset of facial images with potentially noisy expression labels and outputs a more robust and accurate model capable of correctly identifying emotions. Scientists and engineers working on emotional AI, human-computer interaction, or psychological studies involving facial cues would find this beneficial.

No commits in the last 6 months.

Use this if you are training facial expression recognition models and suspect your image labels might contain errors or ambiguities, which can reduce your model's real-world performance.

Not ideal if your dataset is perfectly clean with no label noise, or if your primary interest is in general object classification rather than facial expressions specifically.

facial-expression-recognition emotional-AI computer-vision image-classification machine-learning-robustness
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 9 / 25
Maturity 8 / 25
Community 18 / 25

How are scores calculated?

Stars

92

Forks

17

Language

Python

License

Last pushed

Oct 04, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/zyh-uaiaaaa/Erasing-Attention-Consistency"

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