hhaAndroid/mmdetection-mini

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1100 stars · 190 forks observed · 2026-08-28

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

A minimal, heavily annotated reimplementation of the mmdetection object detection framework, built from scratch for learning purposes. It mirrors mmdetection's structure and implements one-stage and two-stage detectors like RetinaNet, YOLOv3-v5, FCOS, and Faster R-CNN with detailed code explanations.

Use cases

• learn how mmdetection implements object detection algorithms internally
• understand the code details of detectors like YOLOv5, RetinaNet, and Faster R-CNN
• study annotated implementations of one-stage object detectors
• build a simplified object detection framework from scratch to grasp every component
• quickly verify object detection algorithm ideas on VOC/COCO datasets
• add custom features or visualizations to a detection framework without merge conflicts

When to choose

• you want to deeply understand how mmdetection and object detection frameworks work line by line
• you prefer a small, readable codebase with detailed Chinese-language explanations over the full mmdetection
• you want a minimal framework to experiment with detection algorithm modifications

When to avoid

• you need production training with distributed multi-GPU support, which is not supported
• you want the latest features, best performance, or official model zoo of mmdetection
• you need a stable, actively maintained framework for real projects

Facets

library · maturity maintenance

machine-learning deep-learning computer-vision image-processing computer-vision deep-learning machine-learning education tutorials python object-detection mmdetection pytorch learning-project yolo retinanet faster-rcnn annotated-code linux gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="hhaAndroid/mmdetection-mini")