It is our great pleasure to welcome you to the 5th Audio-Visual Emotion recognition Challenge (AVEC 2015), held in conjunction with the ACM Multimedia 2015. This year's challenge and associated workshop continues to push the boundaries of audio-visual emotion recognition. The first AVEC challenge posed the problem of detecting discrete emotion classes on an extremely large set of natural behaviour data. The second AVEC extended this problem to the prediction of continuous valued dimensional affect on the same set of challenging data. In its third edition, we enlarged the problem even further to include the prediction of self-reported severity of depression. The fourth edition of AVEC focused on the study of depression and affect by narrowing down the number of tasks to be used, and enriching the annotation. Finally, this year we've focused the study of affect by including physiology, along with audio-visual data, in the dataset, making the very first emotion recognition challenge that bridges across audio, video and physiological data.
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, video and -- for the first time ever -- physiological emotion recognition communities, to compare the relative merits of the three 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 un-segmented, non-prototypical and non-preselected data. 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 15 submissions from Asia, Europe, Oceania and North America. The programme committee accepted 9 papers in addition to the baseline paper for oral presentation. For the challenge, no less than 48 results submissions were made by 13 teams! We hope that these proceedings will serve as a valuable reference for researchers and developers in the area of audio-visual-physiological emotion recognition and analysis.
We also encourage attendees to attend the keynote presentation. This valuable and insightful talk can and will guide us to a better understanding of the state of the field, and future direction:
AVEC'15 Keynote Talk -- From Facial Expression Analysis to Multimodal Mood Analysis, Pr. Roland Goecke (University of Canberra, Australia)