Face Detection & Age Gender & Expression & Recognition

Overview

FaceLib:

  • use for Detection, Facial Expression, Age & Gender Estimation and Recognition with PyTorch
  • this repository works with CPU and GPU(Cuda)

Installation

  • Clone and install with this command:
    • with pip and automatic installs everything all you need

      • pip install git+https://github.com/sajjjadayobi/FaceLib.git
    • or with cloning the repo and install required packages

      • git clone https://github.com/sajjjadayobi/FaceLib.git
  • you can see the required packages in requirements.txt

How to use:

  • the simplest way is at example_notebook.ipynb
  • for low-level usage check out the following sections
  • if you have an NVIDIA GPU don't change the device param if not use cpu

1. Face Detection: RetinaFace

  • you can use these backbone networks: Resnet50, mobilenet
    • default weights and model is mobilenet and it will be automatically download
  • for more details, you can see the documentation
  • The following example illustrates the ease of use of this package:
 from facelib import FaceDetector
 detector = FaceDetector()
 boxes, scores, landmarks = detector.detect_faces(image)
  • FaceDetection live on your webcam
   from facelib import WebcamFaceDetector
   detector = WebcamFaceDetector()
   detector.run()

WiderFace Validation Performance on a single scale When using Mobilenet for backbone

Style easy medium hard
Pytorch (same parameter with Mxnet) 88.67% 87.09% 80.99%
Pytorch (original image scale) 90.70% 88.16% 73.82%
Mxnet(original image scale) 89.58% 87.11% 69.12%

2. Face Alignment: Similar Transformation

  • always use detect_align it gives you better performance
  • you can use this module like this:
    • detect_align instead of detect_faces
 from facelib import FaceDetector
 detector = FaceDetector()
 faces, boxes, scores, landmarks = detector.detect_align(image)
  • for more details read detect_image function documentation
  • let's see a few examples
Original Aligned & Resized Original Aligned & Resized
image image image image

3. Age & Gender Estimation:

  • I used UTKFace DataSet for Age & Gender Estimation
    • default weights and model is ShufflenetFull and it will be automatically download
  • you can use this module like this:
   from facelib import FaceDetector, AgeGenderEstimator

   face_detector = FaceDetector()
   age_gender_detector = AgeGenderEstimator()

   faces, boxes, scores, landmarks = face_detector.detect_align(image)
   genders, ages = age_gender_detector.detect(faces)
   print(genders, ages)
  • AgeGenderEstimation live on your webcam
   from facelib import WebcamAgeGenderEstimator
   estimator = WebcamAgeGenderEstimator()
   estimator.run()

4. Facial Expression Recognition:

  • Facial Expression Recognition using Residual Masking Network
    • default weights and model is densnet121 and it will be automatically download
  • face size must be (224, 224), you can fix it in FaceDetector init function with face_size=(224, 224)
  from facelib import FaceDetector, EmotionDetector
 
  face_detector = FaceDetector(face_size=(224, 224))
  emotion_detector = EmotionDetector()

  faces, boxes, scores, landmarks = face_detector.detect_align(image)
  list_of_emotions, probab = emotion_detector.detect_emotion(faces)
  print(list_of_emotions)
  • EmotionDetector live on your webcam
   from facelib import WebcamEmotionDetector
   detector = WebcamEmotionDetector()
   detector.run()
  • on my Webcam 🙂

Alt Text

5. Face Recognition: InsightFace

  • This module is a reimplementation of Arcface(paper), or Insightface(Github)

Pretrained Models & Performance

  • IR-SE50
LFW(%) CFP-FF(%) CFP-FP(%) AgeDB-30(%) calfw(%) cplfw(%) vgg2_fp(%)
0.9952 0.9962 0.9504 0.9622 0.9557 0.9107 0.9386
  • Mobilefacenet
LFW(%) CFP-FF(%) CFP-FP(%) AgeDB-30(%) calfw(%) cplfw(%) vgg2_fp(%)
0.9918 0.9891 0.8986 0.9347 0.9402 0.866 0.9100

Prepare the Facebank (For testing over camera, video or image)

  • the faces images you want to detect it save them in this folder:

    Insightface/models/data/facebank/
              ---> person_1/
                  ---> img_1.jpg
                  ---> img_2.jpg
              ---> person_2/
                  ---> img_1.jpg
                  ---> img_2.jpg
    
  • you can save a new preson in facebank with 3 ways:

    • use add_from_webcam: it takes 4 images and saves them on facebank
       from facelib import add_from_webcam
       add_from_webcam(person_name='sajjad')
    • use add_from_folder: it takes a path with some images from just a person
       from facelib import add_from_folder
       add_from_folder(folder_path='./', person_name='sajjad')
    • or add faces manually (just face of a person not image of a person)
      • I don't suggest this

Using

  • default weights and model is mobilenet and it will be automatically download
    import cv2
    from facelib import FaceRecognizer, FaceDetector
    from facelib import update_facebank, load_facebank, special_draw, get_config
 
    conf = get_config()
    detector = FaceDetector()
    face_rec = FaceRecognizer(conf)
    face_rec.model.eval()
    
    # set True when you add someone new 
    update_facebank_for_add_new_person = False
    if update_facebank_for_add_new_person:
        targets, names = update_facebank(conf, face_rec.model, detector)
    else:
        targets, names = load_facebank(conf)

    image = cv2.imread(your_path)
    faces, boxes, scores, landmarks = detector.detect_align(image)
    results, score = face_rec.infer(conf, faces, targets)
    print(names[results.cpu()])
    for idx, bbox in enumerate(boxes):
        special_draw(image, bbox, landmarks[idx], names[results[idx]+1], score[idx])
  • Face Recognition live on your webcam
   from facelib import WebcamVerify
   verifier = WebcamVerify(update=True)
   verifier.run()
  • example of run this code:

image

Reference:

Owner
Sajjad Ayobi
Data Science Lover, a Little Geek
Sajjad Ayobi
CONditionals for Ordinal Regression and classification in PyTorch

CONDOR pytorch implementation for ordinal regression with deep neural networks. Documentation: https://GarrettJenkinson.github.io/condor_pytorch About

7 Jul 25, 2022
A Joint Video and Image Encoder for End-to-End Retrieval

Frozen️ in Time ❄️ ️️️️ ⏳ A Joint Video and Image Encoder for End-to-End Retrieval project page | arXiv | webvid-data Repository containing the code,

225 Dec 25, 2022
A modular, open and non-proprietary toolkit for core robotic functionalities by harnessing deep learning

A modular, open and non-proprietary toolkit for core robotic functionalities by harnessing deep learning Website • About • Installation • Using OpenDR

OpenDR 304 Dec 28, 2022
Machine learning notebooks in different subjects optimized to run in google collaboratory

Notebooks Name Description Category Link Training pix2pix This notebook shows a simple pipeline for training pix2pix on a simple dataset. Most of the

Zaid Alyafeai 363 Dec 06, 2022
Not Suitable for Work (NSFW) classification using deep neural network Caffe models.

Open nsfw model This repo contains code for running Not Suitable for Work (NSFW) classification deep neural network Caffe models. Please refer our blo

Yahoo 5.6k Jan 05, 2023
converts nominal survey data into a numerical value based on a dictionary lookup.

SWAP RATE Converts nominal survey data into a numerical values based on a dictionary lookup. It allows the user to switch nominal scale data from text

Jake Rhodes 1 Jan 18, 2022
POPPY (Physical Optics Propagation in Python) is a Python package that simulates physical optical propagation including diffraction

POPPY: Physical Optics Propagation in Python POPPY (Physical Optics Propagation in Python) is a Python package that simulates physical optical propaga

Space Telescope Science Institute 132 Dec 15, 2022
A Text Attention Network for Spatial Deformation Robust Scene Text Image Super-resolution (CVPR2022)

A Text Attention Network for Spatial Deformation Robust Scene Text Image Super-resolution (CVPR2022) https://arxiv.org/abs/2203.09388 Jianqi Ma, Zheto

MA Jianqi, shiki 104 Jan 05, 2023
Consumer Fairness in Recommender Systems: Contextualizing Definitions and Mitigations

Consumer Fairness in Recommender Systems: Contextualizing Definitions and Mitigations This is the repository for the paper Consumer Fairness in Recomm

7 Nov 30, 2022
This is the dataset for testing the robustness of various VO/VIO methods

KAIST VIO dataset This is the dataset for testing the robustness of various VO/VIO methods You can download the whole dataset on KAIST VIO dataset Ind

1 Sep 01, 2022
DeepMind's software stack for physics-based simulation and Reinforcement Learning environments, using MuJoCo.

dm_control: DeepMind Infrastructure for Physics-Based Simulation. DeepMind's software stack for physics-based simulation and Reinforcement Learning en

DeepMind 3k Dec 31, 2022
Arch-Net: Model Distillation for Architecture Agnostic Model Deployment

Arch-Net: Model Distillation for Architecture Agnostic Model Deployment The official implementation of Arch-Net: Model Distillation for Architecture A

MEGVII Research 22 Jan 05, 2023
A New Approach to Overgenerating and Scoring Abstractive Summaries

We provide the source code for the paper "A New Approach to Overgenerating and Scoring Abstractive Summaries" accepted at NAACL'21. If you find the code useful, please cite the following paper.

Kaiqiang Song 4 Apr 03, 2022
OOD Generalization and Detection (ACL 2020)

Pretrained Transformers Improve Out-of-Distribution Robustness How does pretraining affect out-of-distribution robustness? We create an OOD benchmark

littleRound 57 Jan 09, 2023
Pytorch Implementation of "Diagonal Attention and Style-based GAN for Content-Style disentanglement in image generation and translation" (ICCV 2021)

DiagonalGAN Official Pytorch Implementation of "Diagonal Attention and Style-based GAN for Content-Style Disentanglement in Image Generation and Trans

32 Dec 06, 2022
DrNAS: Dirichlet Neural Architecture Search

This paper proposes a novel differentiable architecture search method by formulating it into a distribution learning problem. We treat the continuously relaxed architecture mixing weight as random va

Xiangning Chen 37 Jan 03, 2023
Optimizing Value-at-Risk and Conditional Value-at-Risk of Black Box Functions with Lacing Values (LV)

BayesOpt-LV Optimizing Value-at-Risk and Conditional Value-at-Risk of Black Box Functions with Lacing Values (LV) About This repository contains the s

1 Nov 11, 2021
NeuralDiff: Segmenting 3D objects that move in egocentric videos

NeuralDiff: Segmenting 3D objects that move in egocentric videos Project Page | Paper + Supplementary | Video About This repository contains the offic

Vadim Tschernezki 14 Dec 05, 2022
CoCosNet v2: Full-Resolution Correspondence Learning for Image Translation

CoCosNet v2: Full-Resolution Correspondence Learning for Image Translation (CVPR 2021, oral presentation) CoCosNet v2: Full-Resolution Correspondence

Microsoft 308 Dec 07, 2022
Some methods for comparing network representations in deep learning and neuroscience.

Generalized Shape Metrics on Neural Representations In neuroscience and in deep learning, quantifying the (dis)similarity of neural representations ac

Alex Williams 45 Dec 27, 2022