starbucks-logo-segmentation
MobileNetV2-UNet semantic segmentation for Starbucks logo detection - 50ms inference with PyTorch Lightning, binary mask output for mobile deployment
- 6
- Python
- MIT
- 12 mo ago
MobileNetV2-UNet semantic segmentation for Starbucks logo detection - 50ms inference with PyTorch Lightning, binary mask output for mobile deployment
Logo Detection of a custom small dataset. The dataset contains logos of 6 famous brands: Nike, Jordans, Adidas, Puma, Kappa, Quicksilver. The model was developed using the Object-Detection-API by Tensorflow