IRIT - SAMOVA

Audio and video analysis


Our team, at a glance

(~ 8 researchers)

We solve problems related to audio-video content processing in three main fields: Automatic content analysis through statistical models and learning. Modelling for classification and retrieval problems. Audio-Video structuring through multi-channel co-phenomena observation.

Applications

  • Intentionality measuring Automatic video surveillance scene indexing new feature vector for HMM-based packet loss concealment (FREESCALE)
  • Speaker age estimation for elderly speech recognition in European Portuguese (MICROSOFT)
  • Multimodal Human Robot Interaction (Aldebaran Robotics) Evaluations for the REPERE challenge : Multimodal person recognition (ELDA)
  • Automatic punctuation recovery (AUTHOT)
  • Automatic video surveillance scene indexing
  • What are the objects lost by frail elderly at home? A pilot study of 60 people (ASA)
  • Speaker age estimation for elderly speech recognition in European Portuguese (MICROSOFT)
  • Automatic tuning of ear protheses (ARCHEAN)
  • Intentionality measuring
  • Multimodal Human Robot Interaction (Aldebaran Robotics)