Repository navigation
ImageGuard
- Based on efficientnet-b3
- Predictive image classification (neutral, porn)
- The model was only trained for two epochs for experimental purposes only (about 78% accuracy)
How to use?
- Create
modelfolder in project - Copy image_guard_v2.pth into the
modelfolder - Start flask service