Performing in the face of the academy: challenges for a creative practitioner within a university

This piece of writing is an excerpt from a keynote talk given at the Symposium on Artistic Research in Borås, Sweden, on 28 November 2014.

Largest Matching Areas for Illumination and Occlusion Robust Face Recognition

In this paper, we introduce a novel approach to face recognition which simultaneously tackles three combined challenges: 1) uneven illumination; 2) partial occlusion; and 3) limited training data. The new approach performs lighting normalization, occlusion de-emphasis and finally face recognition, based on finding the largest matching area (LMA) at each point on the face, as opposed to traditional fixed-size local area-based approaches. Robustness is achieved with novel approaches for feature extraction, LMA-based face image comparison and unseen data modeling. On the extended YaleB and AR face databases for face identification, our method using only a single training image per person, outperforms other methods using a single training image, and matches or exceeds methods which require multiple training images. On the labeled faces in the wild face verification database, our method outperforms comparable unsupervised methods. We also show that the new method performs competitively even when the training images are corrupted.

Face Recognition in the Scrambled Domain via Salience-Aware Ensembles of Many Kernels

With the rapid development of internet-of-things (IoT), face scrambling has been proposed for privacy protection during IoT-targeted image/video distribution. Consequently in these IoT applications, biometric verification needs to be carried out in the scrambled domain, presenting significant challenges in face recognition. Since face models become chaotic signals after scrambling/encryption, a typical solution is to utilize traditional data-driven face recognition algorithms. While chaotic pattern recognition is still a challenging task, in this paper we propose a new ensemble approach – Many-Kernel Random Discriminant Analysis (MK-RDA) to discover discriminative patterns from chaotic signals. We also incorporate a salience-aware strategy into the proposed ensemble method to handle chaotic facial patterns in the scrambled domain, where random selections of features are made on semantic components via salience modelling. In our experiments, the proposed MK-RDA was tested rigorously on three human face datasets: the ORL face dataset, the PIE face dataset and the PUBFIG wild face dataset. The experimental results successfully demonstrate that the proposed scheme can effectively handle chaotic signals and significantly improve the recognition accuracy, making our method a promising candidate for secure biometric verification in emerging IoT applications.

Perceções sociais dos portugueses face aos riscos associados às alterações climáticas

O aquecimento do sistema climático é inequívoco e a influência humana é clara. A continuação da emissão de gases com efeito de estufa irá potenciar os impactes das alterações climáticas, representando um compromisso futuro que se perpetuará por vários séculos. As alterações climáticas não proporcionam uma experiência sensorial direta, embora as variações de temperatura e de precipitação e os extremos climáticos de vários tipos poderem ser experienciados. O cidadão comum não se apercebe do que está a acontecer, a menos que ocorram mudanças significativas, no estado normal do tempo para uma determinada época, na região do mundo onde ele vive. Mesmo para os especialistas, o problema só é cognoscível através de uma vasta rede científica, técnica e institucional. O conhecimento, é portanto, transmitido ao público em geral maioritariamente através de representações dos media sobre o discurso produzido pela comunidade científica. O principal objetivo deste trabalho é averiguar a interligação entre o discurso científico, o discurso mediático e as perceções da população portuguesa na temática das alterações climáticas. A metodologia utilizada, para a prossecução do objetivo, num trabalho que cruza o domínio científico da engenharia do ambiente com o das ciências sociais, encontra-se dividida em 3 fases principais: (1) Uma primeira fase onde se realiza uma análise ao discurso científico de produção nacional, com uma análise de conteúdo aos resumos dos artigos científicos, utilizando 884 resumos da base de dados Scopus de 1975 a 2013 com a palavra-chave ‘Climate Change’; (2) Uma segunda fase onde se analisa o conteúdo de notícias de meios de comunicação social portugueses, aplicando uma análise de conteúdo a 4340 notícias veiculadas por 4 órgãos de comunicação social (Correio da Manhã, Público, RTP e TSF) entre 2004 e 2013, utilizando os respetivos motores de busca online com a palavra-chave ‘Alterações Climáticas’; e, (3) uma terceira fase onde se compila a informação dos estudos existentes sobre a população portuguesa, utilizando os dados dos Eurobarómetros que incluem o tema das alterações climáticas de 1982 a 2014. Seguindo uma linha condutora que inclui o discurso científico, o discurso mediático e as perceções sociais, almejou-se uma abrangência do tema das alterações climáticas, investigando a existência do fenómeno e as respetivas causas, as consequências com a análise dos impactes e dos riscos associados a esses impactes e as soluções através de medidas de mitigação e de adaptação. Nos principais resultados emerge a evidência de que a intensidade de crescimento da produção científica nacional não se traduz num crescimento consistente dos índices de noticiabilidade dos órgãos de comunicação social e desde 2010 que o número de dias por ano, sem notícias sobre alterações climáticas, ultrapassa dos 50%. Em consequência, os níveis de informação da população portuguesa sobre as alterações climáticas são sistematicamente inferiores à média europeia. Em Portugal as taxas de pouco ou nulo conhecimento rondam os dois terços de inquiridos. Não obstante o seu caráter contínuo, para que as alterações climáticas se tornem alvo de interesse dos media é necessário que ocorram reuniões políticas, encontros científicos ou outros acontecimentos. A visibilidade alcançada pelos acontecimentos nacionais é muito fraca e os acontecimentos meteorológicos extremos, não são frequentemente relacionados com o fenómeno das alterações climáticas. No seio da comunidade científica portuguesa existe um claro consenso sobre a existência das alterações climáticas e das suas causas antropogénicas. No discurso mediático português não se verifica enviesamento da informação não sendo surpreendente os baixos níveis de ceticismo dos portugueses. A abordagem aos impactes das alterações climáticas, tanto no discurso científico como no discurso mediático, é robusta e em 2014, cerca de sete em dez portugueses afirma que as alterações climáticas são um problema muito sério. Contudo, Portugal apresenta a proporção mais baixa da Europa de respondentes que percepcionam as alterações climáticas como o problema mais grave que o mundo enfrenta. Para estes resultados poderá contribuir a baixa inclusão de termos relacionados com o risco, tanto no discurso científico como no mediático, não ultrapassando os 20%. Tanto o discurso científico como o discurso mediático não estão direcionados para as soluções (mencionando especificamente medidas de mitigação ou medidas de adaptação). A menção às medidas de mitigação e às medidas de adaptação não ultrapassam os 16% no discurso científico e apresentam valores ainda mais baixos no discurso mediático (13%). Em ambos os discursos existe uma clara preferência pela menção às medidas de adaptação em detrimento das medidas de mitigação. Com valores tão baixos na abordagem às soluções das alterações climáticas não surpreendem os também diminutos níveis de responsabilidade que a população portuguesa atribui a si própria no combate às alterações climáticas.

Editorial: Facing the Other: Novel Theories and Methods in Face Perception Research

The Editorial on the Research Topic: Facing the Other: Novel Theories and Methods in Face Perception Research

Do people have insight into their face recognition abilities?

Diagnosis of developmental or congenital prosopagnosia (CP) involves self-report of everyday face recognition difficulties, which are corroborated with poor performance on behavioural tests. This approach requires accurate self-evaluation. We examine the extent to which typical adults have insight into their face recognition abilities across four studies involving nearly 300 participants. The studies used five tests of face recognition ability: two that tap into the ability to learn and recognise previously unfamiliar faces (the Cambridge Face Memory Test, CFMT, Duchaine & Nakayama, 2006 and a newly devised test based on the CFMT but where the study phases involve watching short movies rather than viewing static faces – the CFMT-Films) and three that tap face matching (Benton Facial Recognition Test, BFRT, Benton, Sivan, Hamsher, Varney, & Spreen, 1983; and two recently devised sequential face matching tests). Self-reported ability was measured with the 15-item Kennerknecht et al. (2008) questionnaire; two single-item questions assessing face recognition ability; and a new 77-item meta-cognition questionnaire). Overall, we find that adults with typical face recognition abilities have only modest insight into their ability to recognise faces on behavioural tests. In a fifth study, we assess self-reported face recognition ability in people with CP and find that some people who expect to perform poorly on behavioural tests of face recognition do indeed perform poorly. However, it is not yet clear whether individuals within this group of poor performers have greater levels of insight (i.e., into their degree of impairment) than those with more typical levels of performance.

Face matching impairment in developmental prosopagnosia

Developmental prosopagnosia (DP) is commonly referred to as ‘face blindness’, a term that implies a perceptual basis to the condition. However, DP presents as a deficit in face recognition and is diagnosed using memory-based tasks. Here, we test face identification ability in six people with DP, who are severely impaired on face memory tasks, using tasks that do not rely on memory. First, we compared DP to control participants on a standardised test of unfamiliar face matching using facial images taken on the same day and under standardised studio conditions (Glasgow Face Matching Test; GFMT). DP participants did not differ from normative accuracy scores on the GFMT. Second, we tested face matching performance on a test created using images that were sourced from the Internet and so vary substantially due to changes in viewing conditions and in a person’s appearance (Local Heroes Test; LHT). DP participants show significantly poorer matching accuracy on the LHT relative to control participants, for both unfamiliar and familiar face matching. Interestingly, this deficit is specific to ‘match’ trials, suggesting that people with DP may have particular difficulty in matching images of the same person that contain natural day-to-day variations in appearance. We discuss these results in the broader context of individual differences in face matching ability.

Multi-scale keypoints in V1 and face detection

End-stopped cells in cortical area V1, which combine out- puts of complex cells tuned to different orientations, serve to detect line and edge crossings (junctions) and points with a large curvature. In this paper we study the importance of the multi-scale keypoint representa- tion, i.e. retinotopic keypoint maps which are tuned to different spatial frequencies (scale or Level-of-Detail). We show that this representation provides important information for Focus-of-Attention (FoA) and object detection. In particular, we show that hierarchically-structured saliency maps for FoA can be obtained, and that combinations over scales in conjunction with spatial symmetries can lead to face detection through grouping operators that deal with keypoints at the eyes, nose and mouth, especially when non-classical receptive field inhibition is employed. Al- though a face detector can be based on feedforward and feedback loops within area V1, such an operator must be embedded into dorsal and ventral data streams to and from higher areas for obtaining translation-, rotation- and scale-invariant face (object) detection.

Face segregation and recognition by cortical multi-scale line and edge coding

Models of visual perception are based on image representations in cortical area V1 and higher areas which contain many cell layers for feature extraction. Basic simple, complex and end-stopped cells provide input for line, edge and keypoint detection. In this paper we present an improved method for multi-scale line/edge detection based on simple and complex cells. We illustrate the line/edge representation for object reconstruction, and we present models for multi-scale face (object) segregation and recognition that can be embedded into feedforward dorsal and ventral data streams (the “what” and “where” subsystems) with feedback streams from higher areas for obtaining translation, rotation and scale invariance.

Face recognition by cortical multi-scale line and edge representations

Empirical studies concerning face recognition suggest that faces may be stored in memory by a few canonical representations. Models of visual perception are based on image representations in cortical area V1 and beyond, which contain many cell layers for feature extraction. Simple, complex and end-stopped cells provide input for line, edge and keypoint detection. Detected events provide a rich, multi-scale object representation, and this representation can be stored in memory in order to identify objects. In this paper, the above context is applied to face recognition. The multi-scale line/edge representation is explored in conjunction with keypoint-based saliency maps for Focus-of-Attention. Recognition rates of up to 96% were achieved by combining frontal and 3/4 views, and recognition was quite robust against partial occlusions.

Face normalization using multi-scale cortical keypoints

Empirical studies concerning face recognition suggest that faces may be stored in memory by a few canonical representations. Models of visual perception are based on image representations in cortical area V1 and beyond, which contain many cell layers for feature extractions. Simple, complex and end-stopped cells tuned to different spatial frequencies (scales) and/or orientations provide input for line, edge and keypoint detection. This yields a rich, multi-scale object representation that can be stored in memory in order to identify objects. The multi-scale, keypoint-based saliency maps for Focus-of-Attention can be explored to obtain face detection and normalization, after which face recognition can be achieved using the line/edge representation. In this paper, we focus only on face normalization, showing that multi-scale keypoints can be used to construct canonical representations of faces in memory.

Ideias e práticas das mães face à amamentação

Dissertação mest., Psicologia da Saúde, Universidade do Algarve, 2007