# Linzaer/Ultra-Light-Fast-Generic-Face-Detector-1MB

💎1MB lightweight face detection model  (1MB轻量级人脸检测模型)

Repository: https://github.com/Linzaer/Ultra-Light-Fast-Generic-Face-Detector-1MB
Canonical: https://ross.abutalabs.com/products/ultra-light-fast-generic-face-detector-1mb
Language: Python
License: MIT
License Family: permissive
Topics: face-detection, arm, inference, mnn, ncnn
Last push: 2023-12-29T15:39:12+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2519, "days_push": 978, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 7542, forks 1548 (observed 2026-08-28T04:10:00.678420+00:00)

## What it is
An ultra-lightweight face detection model (~1MB FP32, ~300KB quantized) designed for edge computing devices, with slim and RFB variants trained on WIDER FACE. It ships pretrained weights plus inference code for PyTorch, ONNX, NCNN, MNN, Caffe, and OpenCV DNN.

## Use cases
- detect faces in images on mobile or embedded devices
- run face detection on ARM boards with limited memory
- export a small face detection model to onnx for deployment
- quantize a face detector to int8 for faster inference
- add face detection to a camera app on android
- run face detection in c++ with ncnn or mnn

## When to choose
- you need a tiny, fast face detector for edge or mobile hardware
- you want pretrained weights with multiple inference framework backends
- model size and flops matter more than state-of-the-art accuracy

## When to avoid
- you need landmark detection, recognition, or attributes beyond face boxes
- you need the highest accuracy on hard faces and can afford larger models
- you need actively maintained code with recent updates

## Facets
- artifact type: library
- maturity: maintenance
- function: computer-vision, machine-learning, image-processing, llm-inference
- domain: computer-vision, machine-learning, embedded-systems, cross-platform
- platform: python, cpp, windows, embedded, cross-platform
- tags: face-detection, lightweight-model, edge-computing, onnx, ncnn, mnn, caffe, int8-quantization, widerface, pretrained-model, linux, macos, android

## Member repositories
- Linzaer/Ultra-Light-Fast-Generic-Face-Detector-1MB (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:00.678420+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 2026-08-29T17:37:25.075901+00:00, confidence not recorded.
  - readme: https://github.com/Linzaer/Ultra-Light-Fast-Generic-Face-Detector-1MB (fetched 2026-08-28T04:10:00.678420+00:00, sha b26775ad1ed1)
- Data as of 2026-08-30T08:39:29.467469+00:00.