A novel classification-selection approach for the self updating of template-based face recognition systems

Abstract

The boosting on the need of security notably increased the amount of possible facial recognition applications, especially due to the success of the Internet of Things (IoT) paradigm. However, although handcrafted and deep learning-inspired facial features reached a significant level of compactness and expressive power, the facial recognition performance still suffers from intra-class variations such as ageing, facial expressions, lighting changes, and pose. These variations cannot be captured in a single acquisition and require multiple acquisitions of long duration, which are expensive and need a high level of collaboration from the users. Among others, self-update algorithms have been proposed in order to mitigate these problems. Self-updating aims to add novel templates to the users' gallery among the inputs submitted during system operations. Consequently, computational complexity and storage space tend to be among the critical requirements of these algorithms. The present paper deals with the above problems by a novel template-based self-update algorithm, able to keep over time the expressive power of a limited set of templates stored in the system database. The rationale behind the proposed approach is in the working hypothesis that a dominating mode characterises the features' distribution given the client. Therefore, the key point is to select the best templates around that mode. We propose two methods, which are tested on systems based on handcrafted features and deep-learning-inspired autoencoders at the state-of-the-art. Three benchmark data sets are used. Experimental results confirm that, by effective and compact feature sets which can support our working hypothesis, the proposed classification-selection approaches overcome the problem of manual updating and, in case, stringent computational requirements.

RLTIR: Activity-Based Interactive Person Identification via Reinforcement Learning Tree

“…Self-adaptive methods can be utilized in recognition systems to improve the performance from the perspective of model structure. For example, B. Freni et al [21] and G. Orrù et al [22] designed self-updating methods for template-based face recognition systems. A. Mhenni et al [23] proposed a novel user-dependent template update strategy for keystroke recognition.…”

On parameterizing higher-order motion for behaviour recognition

“…Gait analysis is one of the main branches of activity recognition. In biometric gait can be obtained at a distance and hard to hide or disguise compare with face, fingerprint or iris [14]- [16].…”

Incremental Learning from Low-labelled Stream Data in Open-Set Video Face Recognition

“…Some approaches are based on keeping a set of dynamic clusters to summarise class distributions and model their evolution overtime [50]. Others use a few labelled data to initialise a set of models, which are afterwards sequentially updated based on pseudo labelled data [27,39,35].…”

“…Although there are propositions of end-to-end deep learning approaches for incremental semi-supervised learning [29,35], their inherent characteristics make them yet unsuitable to operate online with streaming data. Therefore, for this specific context, we propose to combine the good characteristics of a deep feature encoder, which transfers knowledge from the source domain, with an ensemble method able to provide adaptation to the target domain.…”