This project demonstrates car object detection using deep learning techniques. It is designed to train, evaluate, and deploy object detection models like YOLO and EfficientDet. The dataset includes images of cars annotated with bounding boxes.
The goal of this project is to detect cars in images or video streams. It leverages pre-trained models for transfer learning and provides tools for custom training.
- Data preprocessing and augmentation
- YOLOv5 and EfficientDet models
- Training and validation pipelines
- Model evaluation (mAP, precision, recall)
- Real-time inference on new images and videos
The dataset should be organized as follows: