Real-Time Age-Invariant Face Recognition in Videos Using the ScatterNet Inception Hybrid Network (SIHN)

Saurabh Bodhe, Prathamesh Kapse, Amarjot Singh; The IEEE International Conference on Computer Vision (ICCV), 2019, pp. 0-0

Abstract


Face recognition has become a vital component of various safety and security systems with applications in safety and security systems, law enforcement applications, access control etc. Ageing makes face recognition challenging as the facial features evolve over time. In this paper, we propose a ScatterNet Inception Hybrid Network (SIHN) network that learns deep features for age-invariant face recognition. The trained system is evaluated on a separate dataset of 200 videos corresponding to 100 celebrities collected from public sources. These videos contain faces recorded at different locations, scales, rotations, illumination and ages. Experimental results evaluated over 27000 frames show that the proposed method can achieve state-of-the-art performance on both our video dataset as well as the other widely used datasets for age-invariant face datasets such as CACD and FG-NET. The system finds the individuals of interest from the videos in real-time at 18 fps. This research also introduces the Celebrities Video Aging (CVA) dataset used for evaluating the deep network which hopefully may encourage researchers interested in using deep learning for age-invariant face recognition.

Related Material


[pdf]
[bibtex]
@InProceedings{Bodhe_2019_ICCV,
author = {Bodhe, Saurabh and Kapse, Prathamesh and Singh, Amarjot},
title = {Real-Time Age-Invariant Face Recognition in Videos Using the ScatterNet Inception Hybrid Network (SIHN)},
booktitle = {The IEEE International Conference on Computer Vision (ICCV) Workshops},
month = {Oct},
year = {2019}
}