# GeekAlexis/FastMOT

High-performance multiple object tracking based on YOLO, Deep SORT, and KLT 🚀

Repository: https://github.com/GeekAlexis/FastMOT
Canonical: https://ross.abutalabs.com/products/fastmot
Language: Python
License: MIT
License Family: permissive
Topics: jetson, multi-object-tracking, tensorrt, real-time, reid, object-detection, deep-sort, computer-vision, yolov4, ssd, yolov3, people-counter, scaledyolov4, edge-computing, lucas-kanade, video-analysis, deep-learning
Last push: 2024-07-25T06:06:18+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2407, "days_push": 769, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1220, forks 253 (observed 2026-08-28T04:04:01.854815+00:00)

## What it is
FastMOT is a high-performance multiple object tracking system combining YOLO/SSD detection, Deep SORT with OSNet ReID, and KLT optical flow tracking, optimized with TensorRT and Numba. It achieves real-time tracking speeds even on NVIDIA Jetson edge devices and includes camera motion compensation for moving-camera scenes.

## Use cases
- track multiple people in real-time video
- count people crossing a camera view
- run object tracking on Jetson edge devices
- track objects with a moving camera
- re-identify objects that leave and re-enter the frame
- benchmark multi-object tracking on MOT challenge datasets

## When to choose
- you need real-time multi-object tracking on GPU or Jetson hardware
- your camera moves and standard Deep SORT or FairMOT tracking degrades
- you want high tracking accuracy (near state-of-the-art MOTA) with fast inference
- you want to skip detector frames and fill gaps with optical flow for speed

## When to avoid
- you need tracking on CPU-only machines without CUDA GPUs
- you need a maintained project with frequent updates (last release mid-2024, mostly 2021-era news)
- you need non-Python integration or a plug-and-play GUI application
- you track non-person object classes without your own detector weights

## Facets
- artifact type: library
- maturity: maintenance
- function: computer-vision, image-processing, video-processing, machine-learning, deep-learning
- domain: computer-vision, deep-learning, artificial-intelligence
- platform: python
- tags: multiple-object-tracking, yolo, deep-sort, tensorrt, klt-tracker, reid, nvidia-jetson, real-time, people-counting, edge-computing, video, linux, gpu, docker

## Member repositories
- GeekAlexis/FastMOT (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:01.854815+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:15:55.062452+00:00, confidence not recorded.
  - readme: https://github.com/GeekAlexis/FastMOT (fetched 2026-08-28T04:04:01.854815+00:00, sha 5de9c6171230)
- Data as of 2026-08-30T08:39:29.467469+00:00.