shaoshengsong/DeepSORT

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

1003 stars · 219 forks observed · 2026-09-03

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

A C++ implementation of multi-object tracking (MOT) combining YOLOv5 object detection with DeepSORT and ByteTrack trackers. It uses ONNX Runtime for CPU inference of detection and feature-extraction models, avoiding CUDA/TensorFlow dependencies.

Use cases

• track multiple objects in a video with C++
• run yolov5 detection with deepsort tracking
• implement bytetrack multi-object tracking
• deploy object tracking without GPU or cuda
• use custom re-identification feature embeddings for tracking
• process video streams with object detection and tracking

When to choose

• you need C++ multi-object tracking with YOLOv5 detection
• you want CPU-only inference via ONNX Runtime without CUDA dependencies
• you want to choose between DeepSORT and ByteTrack trackers
• you need customizable feature dimensions for re-identification

When to avoid

• you need a Python tracking pipeline
• you require a maintained project with an explicit license
• you need GPU-accelerated tracking out of the box
• you want detectors other than YOLOv5 without writing your own module

Facets

library · maturity maintenance

computer-vision machine-learning image-processing computer-vision deep-learning cpp cross-platform multi-object-tracking deepsort bytetrack yolov5 onnxruntime object-detection kalman-filter video linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="shaoshengsong/DeepSORT")