Face recognition from video

In spite of over two decades of intense research, illumination and pose invariance remain prohibitively challenging aspects of face recognition for most practical applications. The objective of this work is to recognize faces using video sequences both for training and recognition input, in a realistic, unconstrained setup in which lighting, pose and user motion pattern have a wide variability and face images are of low resolution. In particular there are three areas of novelty: (i) we show how a photometric model of image formation can be combined with a statistical model of generic face appearance variation, learnt offline, to generalize in the presence of extreme illumination changes; (ii) we use the smoothness of geodesically local appearance manifold structure and a robust same-identity likelihood to achieve invariance to unseen head poses; and (iii) we introduce an accurate video sequence "reillumination" algorithm to achieve robustness to face motion patterns in video. We describe a fully automatic recognition system based on the proposed method and an extensive evaluation on 171 individuals and over 1300 video sequences with extreme illumination, pose and head motion variation. On this challenging data set our system consistently demonstrated a nearly perfect recognition rate (over 99.7%), significantly outperforming state-of-the-art commercial software and methods from the literature. © Springer-Verlag Berlin Heidelberg 2006.

Natural versus artificial scene classification by ordering discrete fourier power spectra

Holistic representations of natural scenes is an effective and powerful source of information for semantic classification and analysis of arbitrary images. Recently, the frequency domain has been successfully exploited to holistically encode the content of natural scenes in order to obtain a robust representation for scene classification. In this paper, we present a new approach to naturalness classification of scenes using frequency domain. The proposed method is based on the ordering of the Discrete Fourier Power Spectra. Features extracted from this ordering are shown sufficient to build a robust holistic representation for Natural vs. Artificial scene classification. Experiments show that the proposed frequency domain method matches the accuracy of other state-of-the-art solutions. © 2008 Springer Berlin Heidelberg.

Gaussian mixture modeling with Gaussian process latent variable models

Density modeling is notoriously difficult for high dimensional data. One approach to the problem is to search for a lower dimensional manifold which captures the main characteristics of the data. Recently, the Gaussian Process Latent Variable Model (GPLVM) has successfully been used to find low dimensional manifolds in a variety of complex data. The GPLVM consists of a set of points in a low dimensional latent space, and a stochastic map to the observed space. We show how it can be interpreted as a density model in the observed space. However, the GPLVM is not trained as a density model and therefore yields bad density estimates. We propose a new training strategy and obtain improved generalisation performance and better density estimates in comparative evaluations on several benchmark data sets. © 2010 Springer-Verlag.

Corpus design for a unit selection TtS system with application to Bulgarian

In this paper we present the process of designing an efficient speech corpus for the first unit selection speech synthesis system for Bulgarian, along with some significant preliminary results regarding the quality of the resulted system. As the initial corpus is a crucial factor for the quality delivered by the Text-to-Speech system, special effort has been given in designing a complete and efficient corpus for use in a unit selection TTS system. The targeted domain of the TTS system and hence that of the corpus is the news reports, and although it is a restricted one, it is characterized by an unlimited vocabulary. The paper focuses on issues regarding the design of an optimal corpus for such a framework and the ideas on which our approach was based on. A novel multi-stage approach is presented, with special attention given to language and speaker dependent issues, as they affect the entire process. The paper concludes with the presentation of our results and the evaluation experiments, which provide clear evidence of the quality level achieved. © 2011 Springer-Verlag.

Foot contact detection for sprint training

We introduce a new algorithm to automatically identify the time and pixel location of foot contact events in high speed video of sprinters. We use this information to autonomously synchronise and overlay multiple recorded performances to provide feedback to athletes and coaches during their training sessions. The algorithm exploits the variation in speed of different parts of the body during sprinting. We use an array of foreground accumulators to identify short-term static pixels and a temporal analysis of the associated static regions to identify foot contacts. We evaluated the technique using 13 videos of three sprinters. It successfully identifed 55 of the 56 contacts, with a mean localisation error of 1.39±1.05 pixels. Some videos were also seen to produce additional, spurious contacts. We present heuristics to help identify the true contacts. © 2011 Springer-Verlag Berlin Heidelberg.