BWCFace: Open-set Face Recognition using Body-worn Camera

TL;DR

This paper introduces BWCFace, a large dataset for body-worn camera face recognition and evaluates deep learning models for open-set identification, highlighting the importance of fine-tuning for improved accuracy.

Contribution

The work provides a new dataset of 178K facial images from body-worn cameras and assesses CNN architectures with various loss functions for face recognition in this context.

Findings

Maximum 33.89% Rank-1 accuracy with off-the-shelf SENet-50.

Accuracy improves to 99.00% after fine-tuning on BWCFace.

Performance consistent across different body-worn camera sensors.

Abstract

With computer vision reaching an inflection point in the past decade, face recognition technology has become pervasive in policing, intelligence gathering, and consumer applications. Recently, face recognition technology has been deployed on bodyworn cameras to keep officers safe, enabling situational awareness and providing evidence for trial. However, limited academic research has been conducted on this topic using traditional techniques on datasets with small sample size. This paper aims to bridge the gap in the state-of-the-art face recognition using bodyworn cameras (BWC). To this aim, the contribution of this work is two-fold: (1) collection of a dataset called BWCFace consisting of a total of 178K facial images of 132 subjects captured using the body-worn camera in in-door and daylight conditions, and (2) open-set evaluation of the latest deep-learning-based Convolutional Neural Network (CNN) architectures combined with five different loss functions for face identification, on the collected dataset. Experimental results on our BWCFace dataset suggest a maximum of 33.89% Rank-1 accuracy obtained when facial features are extracted using SENet-50 trained on a large scale VGGFace2 facial image dataset. However, performance improved up to a maximum of 99.00% Rank-1 accuracy when pretrained CNN models are fine-tuned on a subset of identities in our BWCFace dataset. Equivalent performances were obtained across body-worn camera sensor models used in existing face datasets. The collected BWCFace dataset and the pretrained/ fine-tuned algorithms are publicly available to promote further research and development in this area. A downloadable link of this dataset and the algorithms is available by contacting the authors.