13/06/2018
Marie Skłodowska-Curie Actions

Computer Vision Centre - The Human Pose Recovery and Behavior Analysis Group

This hosting has expired


  • OFFER DEADLINE
    15/07/2018 17:00 - Europe/Brussels
  • EU RESEARCH FRAMEWORK PROGRAMME
    H2020 / Marie Skłodowska-Curie Actions
  • LOCATION
    Spain, Cerdanyola del Vallès
  • ORGANISATION/COMPANY
    Computer Vision Centre
  • DEPARTMENT
    Research Projects Office

The Human Pose Recovery and Behavior Analysis group (HuPBA) has more 15 years of experience in research and transfer activities related to Computer Vision object and human recognition from multi-modal data sources. The works of the group define part of the state of the art in the field of research on visual data analysis, with special interest in human analysis and affective computing from different sources of information, including RGB, Depth, and Thermal data. Special interest is in Face analysis, Body posture, 3D from 2D human analysis, action/gesture recognition, and emotion and personality profiling. The members of the group have published hundreds of research papers in relevant journals and conferences, several Patents and model utilities, participation in more than 30 research projects, both public and private, both national and international, including FP6, FP7, H2020. One of the main activities of the group is the international organization on computer vision and machine learning competitions, with special emphasis in the analysis of humans and scenes in still images, image sequences, and multi-modal data. (http://sergioescalera.com/http://www.cvc.uab.es/?page_id=223)

Project description: Analysing human faces, body postures and action/gestures in visual data are key topics in Computer Vision. The analysis of humans within the Looking at People field will push the automation of human behaviours for a large range of applications, including HRI, social robotics and affective computing related scenarios, among others. We expect proposals interested in analysing uni/multi-modal human faces and bodies, either static or dynamic in relation to emotional states and personality profiles. Main methodology will include spatio-temporal deep models, and we envision multi-disciplinary impact of the research given its relation to behavioural phycology and psychiatry.