This book constitutes the thoroughly refereed post-proceedings of the Third International Workshop on Machine Learning for Multimodal Interaction, MLMI 2006, held in Bethseda, MD, USA, in May 2006.

The 39 revised full papers presented together with 1 invited paper were carefully selected during two rounds of reviewing and revision. The papers are organized in topical sections on multimodal processing, image and video processing, HCI and applications, discourse and dialogue, speech and audio processing, and NIST meeting recognition evaluation.

This book constitutes the thoroughly refereed post-proceedings of the Second International Workshop on Machine Learning for Multimodal Interaction held in July 2005. The 38 revised full papers presented together with two invited papers were carefully selected during two rounds of reviewing and revision. The papers are organized in topical sections on multimodal processing, HCI and applications, discourse and dialogue, emotion, visual processing, speech and audio processing, and NIST meeting recognition evaluation.

This book contains a selection of refereed papers presented at the 1st Wo- shop on Machine Learning for Multimodal Interaction (MLMI 2004), held at the "Centre du Parc," Martigny, Switzerland, during June 21-23, 2004. The workshop was organized and sponsored jointly by three European projects, - AMI, Augmented Multiparty Interaction, http: //www.amiproject.org - PASCAL, Pattern Analysis, Statistical Modeling and Computational Learning, http: //www.pascal-network.org - M4, Multi-modal Meeting Manager, http: //www.m4project.org as well as the Swiss National Centre of Competence in Research (NCCR): - IM2: Interactive Multimodal Information Management, http: //www.im2.ch MLMI 2004 was thus sponsored by the European Commission and the Swiss National Science Foundation. Given the multiple links between the above projects and several related - search areas, it was decided to organize a joint workshop bringing together - searchers from the di?erent communities working around the common theme of advanced machine learning algorithms for processing and structuring mul- modal human interaction in meetings. The motivation for creating such a forum, which could be perceived as a number of papers from di?erent research dis- plines, evolved from a real need that arose from these projects and the strong motivation of their partners for such a multidisciplinary workshop. This asse- ment was indeed con?rmed by the success of this ?rst MLMI workshop, which attracted more than 200 participants.