YOLOv5 is a PyTorch-based model for vision AI tasks, providing training, inference, and deployment for object detection, instance segmentation, and image classification.
Project overview
The project is recognized for combining speed, accuracy, and simplicity in vision AI tasks, offering multi-input inference and model export to multiple deployment formats.
Project type
Model Development
Deployment
Refer to project documentation
License
AGPL-3.0
Best for
Developers, AI engineers, and researchers who need a PyTorch model for object detection, instance segmentation, and image classification tasks.
Users who need to run inference on diverse inputs such as images, videos, webcams, streams, and directories.
Key capabilities
Runs inference to detect objects on images, videos, webcams, streams, and directories.
Performs instance segmentation on images, including training, validation, and prediction scripts.
Supports image classification tasks.
Allows training YOLOv5 models on custom datasets using GPU configurations.
Exports trained models to deployment formats including TFLite, ONNX, CoreML, and TensorRT.
Limitations and risks
PyTorch Hub inference support for segmentation models might vary.
Getting started
Clone the YOLOv5 repository, install required packages including PyTorch>=1.8, and run inference using a webcam or image.