abewley/sort

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

4373 stars · 1128 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

SORT is a barebones Python implementation of a simple online and realtime multiple object tracking algorithm for 2D video sequences, based on Kalman filtering and data association. It serves as a baseline and testbed for developing future trackers and was the top open-source tracker on the MOT benchmark at publication.

Use cases

• track multiple objects in a video in realtime
• assign consistent identities to detections across frames
• baseline multiple object tracker for research
• run tracking on MOT benchmark sequences
• online tracking using only past and current frames

When to choose

• you need a fast, simple, realtime online tracker as a baseline
• you want a lightweight Kalman-filter-based tracker to extend or compare against
• you're working with MOT benchmark data and provided detections

When to avoid

• you need occlusion handling or re-identification of lost objects (use Deep SORT instead)
• you need a production-grade tracker with permissive licensing (it's GPL-3.0)
• you need built-in object detection (it relies on external detectors)

Facets

library · maturity maintenance

computer-vision image-processing machine-learning computer-vision machine-learning python cross-platform multiple-object-tracking kalman-filter hungarian-algorithm realtime-tracking mot-benchmark research-code video

Member repositories

For agents

markdown · JSON · MCP: product_card(name="abewley/sort")