Proceedings of the 6th International Workshop on Audio/Visual Emotion Challenge

Abstract

It is our great pleasure to welcome you to the 6th Audio-Visual Emotion recognition Challenge -- Depression, Mood, and Emotion (AVEC 2016), held in conjunction with the ACM Multimedia 2016 in Amsterdam, The Netherlands. This year's challenge and associated workshop continues to push the boundaries of audio-visual emotion recognition, and sees a return to the Behaviomedical topic of automatic depression recognition. The first and second AVEC challenges posed the problem of detecting emotions on an extremely large set of natural audio-visual behaviour data. In its third and fourth editions, we extended the problem even further to include the prediction of self-reported severity of depression, which is a frequently occurring mood disorder. The fifth edition included physiological data for emotion prediction. This year we see sub-challenges on two datasets -- RECOLA that was used also in AVEC 2015 and the newly introduced DAIC-WOZ dataset. It is our intention to make it a tradition to repeat each sub-challenge for a second year, while also introducing a new sub-challenge. We hope this encourages more participants to take part and will increase interaction between researchers from closely related research areas. The mission of AVEC challenge and workshop series is to provide a common benchmark test set for individual multimodal information processing and to bring together the audio and video emotion recognition communities, to compare the relative merits of the two approaches to emotion recognition under well-defined and strictly comparable conditions and establish to what extent fusion of the approaches is possible and beneficial. A second motivation is the need to advance emotion recognition systems to be able to deal with naturalistic behaviour in large volumes of unsegmented, non-prototypical and non-preselected. As you will see, these goals have been reached with the selection of this year's data and the challenge contributions. The call for participation attracted 14 submissions from Asia, Europe, and North America. The programme committee accepted 12 papers in addition to the baseline paper for oral presentation. For the depression sub-challenge we received submissions by 7 teams, and for the emotion subchallenge a record 13 teams submitted results! We hope that these proceedings will serve as a valuable reference for researchers and developers in the area of audio-visual emotion recognition and depression analysis. We also encourage attendees to attend the keynote presentations. This valuable and insightful talk can and will guide us to a better understanding of the state of the field, and future directions: Personality and Emotion Analysis, Dr Hatice Gunes (University of Cambridge, UK)