CONF Favre_ICMI2008_2008/IDIAP Role Recognition in Multiparty Recordings using Social Affiliation Networks and Discrete Distributions Favre, Sarah Salamin, Hugues Dines, John Vinciarelli, Alessandro EXTERNAL http://publications.idiap.ch/attachments/papers/2009/Favre_ICMI2008_2008.pdf PUBLIC http://publications.idiap.ch/index.php/publications/showcite/sfavre:rr08-64 Related documents International Conference on Multimodal Interfaces Chania, Greece 2008 October 2008 This paper presents an approach for the recognition of roles in multiparty recordings. The approach includes two major stages: extraction of Social Affiliation Networks (speaker diarization and representation of people in terms of their social interactions,',','), and role recognition (application of discrete probability distributions to map people into roles). The experiments are performed over several corpora, including broadcast data and meeting recordings, for a total of roughly 90 hours of material. The results are satisfactory for the broadcast data (around 80 percent of the data time correctly labeled in terms of role,',','), while they still must be improved in the case of the meeting recordings (around 45 percent of the data time correctly labeled). In both cases, the approach outperforms significantly chance. REPORT sfavre:rr08-64/IDIAP Role Recognition in Multiparty Recordings using Social Affiliation Networks and Discrete Distributions Favre, Sarah Salamin, Hugues Vinciarelli, Alessandro EXTERNAL http://publications.idiap.ch/attachments/reports/2008/sfavre-idiap-rr-08-64.pdf PUBLIC Idiap-RR-64-2008 2008 IDIAP To appear in Proceedings of ICMI International Conference on Multimodal Interfaces (2008) This paper presents an approach for the recognition of roles in multiparty recordings. The approach includes two major stages: extraction of Social Affiliation Networks (speaker diarization and representation of people in terms of their social interactions,',','), and role recognition (application of discrete probability distributions to map people into roles). The experiments are performed over several corpora, including broadcast data and meeting recordings, for a total of roughly 90 hours of material. The results are satisfactory for the broadcast data (around 80 percent of the data time correctly labeled in terms of role,',','), while they still must be improved in the case of the meeting recordings (around 45 percent of the data time correctly labeled). In both cases, the approach outperforms significantly chance.