Hugging Face Model Integrity Monitor
8.2
A monitoring and alerting system that proactively identifies and flags Hugging Face models exhibiting malicious behavior, specifically targeting the creation of non-consensual deepfakes. It employs anomaly detection algorithms to detect unusual image editing patterns and automatically alerts administrators.
160h
mvp estimate
8.2
viability grade
24
views
technology stack
Python
PostgreSQL
Medium
inspired by
Hugging Face used to easily undress women and children