⚡️ SUPER HIGH SPEED RECOGNITION
⚡️ USEFUL MANY METHODS
⚡️ RICH AND COMPREHENSIVE DOCUMENTATION
FACE01 -- LET'S START !
# result
Audrey Hepburn
Anti spoof real
Anti spoof score 100.0 %
similarity 99.1%
coordinate (123, 390, 334, 179)
time 2022,08,09,04,19,35,552949
output output/Audrey Hepburn_2022,08,09,04,19,35,556237_0.39.png
-------
📖 TOC
✨ FACE01 is a face recognition library that integrates various functions and can be called from Python.
- 🎉 まばたき検知を実装しました。簡易なりすまし防止機能として実装可能です。
- 🎉 JAPANESE FACE v1 is now available !
JAPANESE FACE v1
は日本人の顔認証に特化したモデルです。- コード中では
EfficientNetV2 Arcface Model
と表現されています。
- コード中では
Real-time face recognition
is possible from face datas of more than 10,000 people- Super high-speed face coordinate output function
- Face image saving function with date and time information
- You can set to modify output frame image
- Centralized management of functions by configuration file
- You can choose input protocol ex. RTSP, HTTP and USB
- You can use many function for
face-recognition
andImage-processing
(See Useful FACE01 library) - ...and many others!
- このリポジトリが提供するファイルは、無料でお使いいただけます。 教育機関でご利用の場合、ソースコードを研究・教育にご利用できます。 詳しくは日本のAI教育を支援する、顔認識ライブラリ
FACE01
の提供についてをご覧ください。- 商用利用にあたっては別途ライセンスが必要です。
- こちらのサンプルは2024年2月までご使用になれます。これ以降にこのサンプルを試用いただく場合にはご連絡をください。
Setting up your FACE01 develop environment is really easy !
wget https://raw.githubusercontent.com/yKesamaru/FACE01_SAMPLE/master/INSTALL_FACE01.sh
chmod +x INSTALL_FACE01.sh
bash -c ./INSTALL_FACE01.sh
See here.
🐳 The easiest way to use Docker is to pull the image.
See here.
If you cannot use Docker by any means, please refer to here.
There are some example files in the example folder.
Let's try step-by-step examples.
See here.
If you want to see the exhaustive document, see here.
- Highly flexible, inheritable and easy-to-use configuration file: config.ini See here.
- 🔖 v2.1.05
- Add function to detect eye blinks.
ℹ️
This repository contains FACE01 SAMPLE for UBUNTU 20.04.
If you are a Windows user, please use this on Docker.
This sample can be used until December 2023.
📄 I would like to acknowledgments those who have published such wonderful libraries and models.
- dlib / davisking
- face_recognition / ageitgey
- mediapipe / google
- open_model_zoo / openvinotoolkit
- light-weight-face-anti-spoofing / kprokofi
- openvino2tensorflow / Katsuya Hyodo (PINTO0309)
- PINTO_model_zoo / Katsuya Hyodo (PINTO0309)
- FaceDetection-Anti-Spoof-Demo / KazuhitoTakahashi (Kazuhito00)
- Some images from Pakutaso, pixabay
- Deep Face Recognition A Survey
- EfficientNetV2: Smaller Models and Faster Training
- ArcFace: Additive Angular Margin Loss for Deep
- MobileFaceNets: Efficient CNNs for Accurate Real-Time Face Verification on Mobile Devices
- Dlib Python API
- Pytorch documentation and Python API
- ONNX documentation
- 教育と著作権: 水谷正大 著, 大東文化大学 (2021)