MLMI stands for Machine Learning for Multimodal Interaction. It is an annual workshop bringing together researchers working on advanced machine learning algorithms applied to multimodal human-human and human-computer interaction. The first edition was in 2004 featuring a keynote from Bill Buxton. Every year it was recorded and put online.
Klewel was part of the capturing. The conference was held in Utretch – not far from Amsterdam, The Neitherlands. The MLMI workshop 2008 videos are available both on the project site as well as on Klewel portal.