Repolode
Submit
serengil

serengil/deepface

A Python library for face recognition and facial attribute analysis, covering age, gender, emotion and race, with multiple pluggable models and detectors.

View on GitHubWebsite
Stars
23,653
Forks
3,210
Open issues
4
Last commit
2 days ago

What is it?

DeepFace is a Python framework for face recognition and facial attribute analysis. It wraps several face recognition models, including VGG-Face, FaceNet, ArcFace, OpenFace and SFace, and a range of face detectors such as OpenCV, RetinaFace, MTCNN, MediaPipe and YOLO variants. Developers can call high-level functions like verify, find, analyze and represent in a single line, while the library handles detection, alignment, normalization and vector comparison in the background.

Beyond one-to-one verification and folder-based search, DeepFace supports database-backed registration and search through backends such as PostgreSQL, pgvector, MongoDB, Redis and Qdrant, with optional approximate nearest neighbor search. It also provides real-time webcam streaming, anti-spoofing checks, and a REST and gRPC API that can run directly or in Docker. The default inference engine is ONNX Runtime, with TensorFlow and PyTorch available as optional extras.

Key features

  • 01Face verification and 1:N face search over folders or databases.
  • 02Age, gender, emotion and race attribute analysis.
  • 03Multiple recognition models and detector backends that can be switched by argument.
  • 04Real-time webcam streaming with recognition and attribute analysis.
  • 05Optional anti-spoofing check to detect fake faces.
  • 06REST and gRPC API servers, with Docker support.

Use cases

  • Building a face login or identity check feature in a Python application.
  • Searching a photo collection for a specific person.
  • Running face analysis as a microservice called from a web or mobile app.
  • Prototyping computer vision projects that need face embeddings.

Quick start

  1. 1

    Install the library from PyPI (ONNX Runtime backend by default)

    pip install deepface
  2. 2

    Import the library in Python

    from deepface import DeepFace
  3. 3

    Verify whether two face images show the same person

    result: dict = DeepFace.verify(img1_path = "img1.jpg", img2_path = "img2.jpg")

Things to know

  • Face recognition and attribute analysis raise privacy and ethical concerns, so check the legal rules where you deploy it.
  • Full TensorFlow or PyTorch backends add one to several GB of dependencies.

In collections

Computer vision libraries & tools

Topics

#face-recognition#computer-vision#deep-learning#facial-analysis#python#embeddings#vector-search#age-prediction#arcface#deepface

FAQ

What is deepface?
A Python library for face recognition and facial attribute analysis, covering age, gender, emotion and race, with multiple pluggable models and detectors. DeepFace is a Python framework for face recognition and facial attribute analysis. It wraps several face recognition models, including VGG-Face, FaceNet, ArcFace, OpenFace and SFace, and a range of face detectors such as OpenCV, RetinaFace, MTCNN, MediaPipe and YOLO variants. Developers can call high-level functions like verify, find, analyze and represent in a single line, while the library handles detection, alignment, normalization and vector comparison in the background.
Is deepface free to use? What license does it have?
deepface is open source under the MIT license. Check the license terms before using it in commercial products.
Is deepface still maintained?
Actively maintained. The last commit to deepface was on Nov 5, 2026.
What can I use deepface for?
Building a face login or identity check feature in a Python application. Searching a photo collection for a specific person. Running face analysis as a microservice called from a web or mobile app. Prototyping computer vision projects that need face embeddings.
How do I install deepface?
Install the library from PyPI (ONNX Runtime backend by default): pip install deepface Import the library in Python: from deepface import DeepFace Verify whether two face images show the same person: result: dict = DeepFace.verify(img1_path = "img1.jpg", img2_path = "img2.jpg")
Where can I get deepface?
The source code and installation instructions are on GitHub: https://github.com/serengil/deepface

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects