Novel Framework: Face Feature Selection Algorithm for Neonatal Facial and Related Attributes Recognition

Muhammad Awais,Xi Long,Saadullah Farooq Abbasi,Bin Yin,Wei Chen,Ronald M Aarts,Chunmei Lu,Anum Nawaz,Chen Chen,Laishuan Wang,Saeed Akbarzadeh,Linkai Tao

doi:10.1109/access.2020.2982865

Abstract

In recent times, with the advancement of digital imaging, automatic facial recognition has been intensively studied for adults, while less for neonates. Due to the miniature facial structure and facial attributes, newborn facial recognition remains a challenging area. In this paper, an automatic video-based Neonatal Face Attributes Recognition (NFAR) approach in a hierarchical framework is proposed by coalescing the intensity-based method, pose estimation, and novel dedicated neonatal Face Feature Selection (FFS) algorithm. The intensity-based method is used for face detection, followed by the facial pose estimation algorithm and FFS are dedicated to neonatal pose and face feature recognition, respectively. In this study, video-data of 19 neonates' were collected from the Children's Hospital affiliated to Fudan University, Shanghai, to evaluate the proposed NFAR approach. The results show promising performance to detect the neonatal face, pose estimation (-45°, 45°), and facial features (nose, mouth, and eyes) recognition. The NFAR approach exhibits a sensitivity, accuracy, and specificity of 98.7%, 98.5%, and, 95.7% respectively, for the newborn babies at the frontal (0°) facial region. The neonatal face and its attributes recognition can be expected to detect neonate's medical abnormalities unobtrusively by examining the variation in newborn facial texture pattern.

Highlights

Face is one of the most unique and distinct attributes of a human, which can convey relevant information such as age, gender, emotion, etc
Input video frames acts as an input for intensity-based detection to detect the neonatal facial region, the detected face region is used by pose estimation to estimate the pose and facial area, respectively
RESULTS we investigate and evaluate the performance of existing face detection algorithm methods and our proposed algorithms named ‘‘Neonatal Face Attributes Recognition (NFAR).’’ for the infant’s face, and its attributes recognition

Summary

Introduction

Face is one of the most unique and distinct attributes of a human, which can convey relevant information such as age, gender, emotion, etc. The neonatal facial structure contains approximately 10,000 nerves along with facial attributes that are still immature [1], [2]. Neonate’s facial rationalization has emerged as a spry area of research for various applications, e.g., baby swapping [7], baby abduction [8], neonatal pain and sedation scale via change in face pattern [9], infant pain scale measurement using facial expression moment [10], crying relating to variates in facial expression [11]–[13], etc. Due to the non-maturity in neonate’s facial features, expressions, and random changes in their facial pattern and pose [3], [14], neonatal face and its attributes

Objectives

Methods

Results

Discussion

Conclusion