Mobile based Human Identification using Forehead Creases: Application and Assessment under COVID-19 Masked Face Scenarios

This repository contains the code implementation for our paper, which was accepted at WACV 2022 conference.

Table of Contents

• Abstract
• Usage
• Results
• Database
• Reference
• Attribution
• Contact

Abstract

In the COVID-19 situation, face masks have become an essential part of our daily life. As mask occludes most prominent facial characteristics, it brings new challenges to the existing facial recognition systems. This paper presents an idea to consider forehead creases (under surprise facial expression) as a new biometric modality to authenticate mask-wearing faces. The forehead biometrics utilizes the creases and textural skin patterns appearing due to voluntary contraction of the forehead region as features. The proposed framework is an efficient and generalizable deep learning framework for forehead recognition. Face-selfie images are collected using smartphone's frontal camera in an unconstrained environment with various indoor/outdoor realistic environments. Acquired forehead images are first subjected to a segmentation model that results in rectangular Region Of Interest (ROI's). A set of convolutional feature maps are subsequently obtained using a backbone network. The primary embeddings are enriched using a dual attention network (DANet) to induce discriminative feature learning. The attention-empowered embeddings are then optimized using Large Margin Cosine Loss (LMCL) followed by Focal Loss to update weights for inducting robust training and better feature discriminating capabilities. Our system is end-to-end and few-shot; thus, it is very efficient in memory requirements and recognition rate. Besides, we present a forehead image dataset (BITS-IITMandi-ForeheadCreases Images Database) that has been recorded in two sessions from 247 subjects containing a total of 4,964 selfie-face mask images. To the best of our knowledge, this is the first to date mobile-based forehead dataset and is being made available along with the mobile application in the public domain. The proposed system has achieved high performance results in both closed-set, i.e., CRR of 99.08% and EER of 0.44% and open-set matching, i.e., CRR: 97.84%, EER: 12.40% which justifies the significance of using forehead as a biometric modality.

Usage

• Clone this repository to your local machine.

$ git clone https://github.com/rohit901/Forehead-Creases.git

• Install the following required dependencies/prerequisite software.
  1. Anaconda: https://docs.anaconda.com/anaconda/install/index.html
  2. Pytorch: https://pytorch.org
  3. Jupyter Lab/Jupyter Notebook: https://jupyter.org/install
  4. Android Studio (to install and run the data collection application): https://developer.android.com/studio
  5. Any other missing dependencies can be installed by running pip install <missing package name>
• Code for the data-collection Android application can be found in folder named Data_Collection_App. Open this folder in Android Studio to use the application. To use this application, you would also require a backend-server with your backend API hosted. Change the API URL in the code accordingly.
• Network code can be run directly in or by downloading Main_Network_Code.ipynb file and opening it locally using jupyter notebook.

Results

	Subject Count	Matching Strategy	CRR(%)	EER(%)	DI
Model V1	75	Open Set	96.08	17.50	1.44
Model V2	75	Open Set	97.84	12.40	1.66
Model V1	247	Closed Set	97.22	4.46	1.91
Model V2	247	Closed Set	99.08	0.44	2.79

Please refer to our paper for the definition of Model V1 and Model V2.

Database

You can access the BITS-IITMandi-ForeheadCreases Images Database by visiting this website and filling up the database request form. After filling the form you will receive an email acknolwedgement of your submitted responses. Allow upto one week to receive update back from us. Please cite this paper in your work if you use the database.

Reference

Bharadwaj, R., Jaswal, G., Nigam, A., & Tiwari, K. (2022). Mobile based human identification using forehead creases: application and assessment under COVID-19 masked face scenarios. 2022 IEEE Winter Conference on Applications of Computer Vision (WACV), accepted for publication.

Attribution

Camera2API and CameraX library from Android has been used to develop the data-collection Android application. Various other open-source libraries used in this project have been credited appropriately.

Contact

For any queries regarding code or implementation, you can reach out to me on the following places:

• Twitter: @rohit901
• LinkedIn: @rohit901
• Website: rohit901.github.io
• Mail: [email protected]

My co-authors/supervisors can be reached at the following places:

• Gaurav Jaswal: [email protected]
• Aditya Nigam: [email protected]
• Kamlesh Tiwari: [email protected]