-
Pre-training strategies and datasets for facial representation learning
Authors:
Adrian Bulat,
Shiyang Cheng,
Jing Yang,
Andrew Garbett,
Enrique Sanchez,
Georgios Tzimiropoulos
Abstract:
What is the best way to learn a universal face representation? Recent work on Deep Learning in the area of face analysis has focused on supervised learning for specific tasks of interest (e.g. face recognition, facial landmark localization etc.) but has overlooked the overarching question of how to find a facial representation that can be readily adapted to several facial analysis tasks and datase…
▽ More
What is the best way to learn a universal face representation? Recent work on Deep Learning in the area of face analysis has focused on supervised learning for specific tasks of interest (e.g. face recognition, facial landmark localization etc.) but has overlooked the overarching question of how to find a facial representation that can be readily adapted to several facial analysis tasks and datasets. To this end, we make the following 4 contributions: (a) we introduce, for the first time, a comprehensive evaluation benchmark for facial representation learning consisting of 5 important face analysis tasks. (b) We systematically investigate two ways of large-scale representation learning applied to faces: supervised and unsupervised pre-training. Importantly, we focus our evaluations on the case of few-shot facial learning. (c) We investigate important properties of the training datasets including their size and quality (labelled, unlabelled or even uncurated). (d) To draw our conclusions, we conducted a very large number of experiments. Our main two findings are: (1) Unsupervised pre-training on completely in-the-wild, uncurated data provides consistent and, in some cases, significant accuracy improvements for all facial tasks considered. (2) Many existing facial video datasets seem to have a large amount of redundancy. We will release code, and pre-trained models to facilitate future research.
△ Less
Submitted 20 July, 2022; v1 submitted 30 March, 2021;
originally announced March 2021.
-
CryptoCam: Privacy Conscious Open Circuit Television
Authors:
Gerard Wilkinson,
Dan Jackson,
Andrew Garbett,
Reuben Kirkham,
Kyle Montague
Abstract:
The prevalence of Closed Circuit Television (CCTV) in today's society has given rise to an inherent asymmetry of control between the watchers and the watched. A sense of unease relating to the unobservable observer (operator) often leads to a lack of trust in the camera and its purpose, despite security cameras generally being present as a protective device. In this paper, we detail our concept of…
▽ More
The prevalence of Closed Circuit Television (CCTV) in today's society has given rise to an inherent asymmetry of control between the watchers and the watched. A sense of unease relating to the unobservable observer (operator) often leads to a lack of trust in the camera and its purpose, despite security cameras generally being present as a protective device. In this paper, we detail our concept of Open Circuit Television and prototype CryptoCam, a novel system for secure sharing of video footage to individuals and potential subjects nearby. Utilizing point-of-capture encryption and wireless transfer of time-based access keys for footage, we have developed a system to encourage a more open approach to information sharing and consumption. Detailing concerns highlighted in existing literature we formalize our over-arching concept into a framework called Open Circuit Television (OCTV). Through CryptoCam we hope to address this asymmetry of control by providing subjects with data equity, discoverability and oversight.
△ Less
Submitted 18 April, 2020;
originally announced April 2020.
-
A Network Epidemic Model for Online Community Commissioning Data
Authors:
Clement Lee,
Andrew Garbett,
Darren J. Wilkinson
Abstract:
A statistical model assuming a preferential attachment network, which is generated by adding nodes sequentially according to a few simple rules, usually describes real-life networks better than a model assuming, for example, a Bernoulli random graph, in which any two nodes have the same probability of being connected, does. Therefore, to study the propogation of "infection" across a social network…
▽ More
A statistical model assuming a preferential attachment network, which is generated by adding nodes sequentially according to a few simple rules, usually describes real-life networks better than a model assuming, for example, a Bernoulli random graph, in which any two nodes have the same probability of being connected, does. Therefore, to study the propogation of "infection" across a social network, we propose a network epidemic model by combining a stochastic epidemic model and a preferential attachment model. A simulation study based on the subsequent Markov Chain Monte Carlo algorithm reveals an identifiability issue with the model parameters. Finally, the network epidemic model is applied to a set of online commissioning data.
△ Less
Submitted 20 July, 2017; v1 submitted 24 February, 2017;
originally announced February 2017.