📸 Automatically detects and crops faces from batches of pictures.
-
Updated
Oct 19, 2026 - Python
📸 Automatically detects and crops faces from batches of pictures.
YuNetのPythonでのONNX、TensorFlow-Lite推論サンプル
Open-source face login and walk-away auto-lock for Windows 10/11. Credential Provider + DeepFace (ArcFace) + YuNet + MiniFASNet liveness. Tray-managed, 12-language UI. A self-hosted alternative to Windows Hello / Howdy for Windows.
Detecting and anonymizing human faces in images using deep learning models (YuNet) and OpenCV.
GPU-accelerated face detection, cropping, enhancement, and batch export tools in Rust.
A real-time face and object detection with gender classification and virtual glasses overlay.
参考图未通过人像审核?本地遮眼预处理 · Seedance/即梦/豆包。CPU 小模型,零云端视觉费用。Demo: face.83zi.com
Pre-release likeness-risk screening for AI short-drama character images. Finds public-figure lookalikes with traceable references; human review required.
Real-time Multi-face Tracking in TouchDesigner with OpenCV & YuNet
Cluster Face sorts images based on the faces in it.
YUNet evaluation on WIDERFace dataset
Yunet is one of the fastest (maybe lightest) face detection in the world with only 75M params but achieve pretty well performance. There are some thing we can improve on top of it and this repo is to do so.
YUNet implementation using PyTorch
本地离线人脸识别登录系统:Python + OpenCV (YuNet/SFace) + Flask + SQLite,口令 + 人脸 1:1 验证 + 随机动作挑战反欺骗
Convert YuNet models for use on RV1106 NPU
My learning on the AI field also working on several Google Colab Notebooks
TraceMark is an end-to-end face verification pipeline that detects and encodes a face, performs an automated Google Lens reverse-image search through SerpApi, retrieves matching web results, records the discovered data on a local blockchain, and re-verifies its integrity.
Real-time face detection & recognition in C++ with OpenCV (YuNet + SFace), webcam enrollment, and live FPS display.
摄像头隐私屏:本地人脸识别判断主人在否,非主人入镜自动模糊全屏。Windows WPF 桌面工具,画面与特征全部本机处理,数据不出本机。
To associate your repository with the yunet topic, visit your repo's landing page and select "manage topics."