yolov5

Ultralytics YOLOv5 is a widely used open-source computer vision framework built on PyTorch. It provides pretrained and trainable models for object detection, instance segmentation, and image classification tasks. Key capabilities include: - Inference via PyTorch Hub or the detect.py script, supporting images, videos, webcams, RTSP streams, and screen captures - Training on custom datasets with support for single and multi-GPU setups - Model export to ONNX, TensorRT, CoreML, TFLite, and other deployment formats - Pretrained checkpoints available in nano (n), small (s), medium (m), large (l), and extra-large (x) sizes - Additional tutorials covering transfer learning, test-time augmentation, model pruning, hyperparameter evolution, and hardware deployment (including NVIDIA Jetson) - Integrations with experiment tracking platforms such as ClearML, Weights & Biases, and Comet The project requires Python 3.8+ and PyTorch 1.8+. Models and datasets are downloaded automatically from releases. Training on the COCO dataset is demonstrated with example commands, and benchmark tables report mAP scores and inference speed across model variants.