Éditeur(s) : HAL CCSD Institute of Electrical and Electronics Engineers
Résumé : International audience
This paper presents our response to the first interna- tional challenge on Facial Emotion Recognition and Analysis. We propose to combine different types of features to automatically detect Action Units in facial images. We use one multi-kernel SVM for each Action Unit we want to detect. The first kernel matrix is computed using Local Gabor Binary Pattern histograms and a histogram intersection kernel. The second kernel matrix is computed from AAM coefficients and an RBF kernel. During the training step, we combine these two types of features using the recently proposed SimpleMKL algorithm. SVM outputs are then averaged to exploit temporal information in the sequence. To eval- uate our system, we perform deep experimentations on several key issues: influence of features and kernel function in histogram- based SVM approaches, influence of spatially-independent in- formation versus geometric local appearance information and benefits of combining both, sensitivity to training data and interest of temporal context adaptation. We also compare our results to those of the other participants and try to explain why our method had the best performance during the FERA challenge.
ISSN: 1083-4419

hal-00731864
http://hal.upmc.fr/hal-00731864
http://hal.upmc.fr/hal-00731864/document
http://hal.upmc.fr/hal-00731864/file/senechal12tsmc_preprint.pdf
DOI : 10.1109/TSMCB.2012.2193567

Éditeur(s) : HAL CCSD
Résumé : International audience
This paper investigates the narrative challenges and opportunities that contemporary virology offers to popular science and to fiction. The study of viruses reveals extremely complex relations, playing out over different physical and temporal scales. Viruses constitute a heterogeneous population whose interactions with living organisms occur on multiple levels, and whose numbers are vastly superior to human and even bacterial populations on earth. The resulting challenges for narration can be grouped into three main formal questions. The first is the problem of scale, since the numbers and sizes of virus and host populations are vastly different, as are the temporalities over which they transform and survive. The second is the complexity of actants, since viruses are involved in relations where destruction and mutual benefit can overlap, and genetic transformations are played out over many generations. The third is the choice of images through which to convey their behaviours: many existing metaphors in microbiology, such as “chimeras”, “parasites” or “kill-the-winner dynamics”, carry the narrative potential of a mythological and/or theatrical imagination.These questions will first be examined as challenges encountered in the narration of biology for vulgarization purposes, and effects on narrative forms will be outlined: shifts in focalization in order to create multiscalar perspectives; unstable actantial models as expressions of complex relations; and the use or resistance to narrative structures implied by several key metaphors in virological discourse. These formal characteristics will then be presented through two case studies drawn from contemporary fiction. In his “scientific fable” Conflits Intérieurs (2015), microbiologist Eric Bapteste creates a multiscalar narrative by using humans and viruses as both protagonists and narrators in a battle of ideas. In the graphic novel aâma (2011-2014), Frederik Peeters explores the ambiguous potential, both symbiotic and threatening, of a meta-organism uniting humans with microscopic agents. Both these novels play with scale, metaphor and actants so as to question the fundaments of human identity. They posit the act of narration as a defining human attribute, and consequently use the narrative voice as an arena in which we observe the weakening of the notion of the subject by microbiology.
“Narrative, Cognition & Science” Symposium
Erlangen, Germany

hal-01434744
https://hal.archives-ouvertes.fr/hal-01434744