Deep Learning architectures for Computer Vision

This repository contains various Deep Learning models and utility functions implemented from scratch for different Computer Vision tasks using Keras 3 and TensorFlow.

Install requirements

pip install -r requirements.txt

Available Models

Currently, the following models are included in this repository:

• ViT (Vision Transformer)
• U-Net (U-Net for image segmentation)
• FOMO (Fewer Objects, More Objects)
• EfficientNet (The well known EfficientNet backbone)
• ResNet (Residual Networks)