roboflow/supervision

Supervision is a Python library that provides a comprehensive set of tools for processing and analyzing computer vision data, particularly object detection and tracking results. It offers utilities for visualizing detections and key points on images and videos, managing dataset formats, and computing evaluation metrics like mAP and precision/recall. The system also includes example scripts demonstrating practical applications such as people counting, heatmaps, and traffic analysis using various model backends.

The project underwent a significant modernization of its build system and development tooling, migrating to pyproject.toml and establishing a robust testing infrastructure. Concurrently, the codebase was cleaned up by removing obsolete modules, while new examples were added to demonstrate object tracking, heatmaps, and dwell time analysis using various model backends.

13 changes

This period focused on establishing a comprehensive public API for the supervision package, introducing new modules for metrics, key points, and drawing utilities. The work included refactoring the codebase into a modular structure with a new base class for annotators and a memory-efficient CompactMask implementation. Extensive test coverage was added to validate the new features and ensure backward compatibility.

33 changes

roboflow/supervision was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point the surveyor at a repository you know and see whether you agree with it.