The Audio/Visual Emotion Challenge and Workshop (AVEC 2014) `3D Dimensional Affect and Depression’ will be the fourth competition event aimed at comparison of multimedia processing and machine learning methods for automatic audio, visual and audio-visual emotion analysis, with all participants competing under strictly the same conditions.

The AVEC 2014 edition features besides emotion recognition the serious and important automatic estimation of levels of depression. Adding objective measures to what is otherwise an entirely subjective process of diagnosing and monitoring depression promises to be an invaluable support to the mental health profession besides the usage in media retrieval systems.

The goal of the Challenge 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 for multimedia retrieval to a level where behaviomedical systems are able to deal with large volumes of non-prototypical naturalistic behaviour in reaction to known stimuli, as this is exactly the type of data that diagnostic tools and other applications would have to face in the real world.

We are calling for teams to participate in two Sub-Challenges: fully-continuous emotion detection from audio, from video, or from audio-visual information, and estimation of one depression level indicator. As benchmarking database an enhanced subset of the AVEC 2013 audio-visual depression corpus of task-specific human-computer interactions will be used. Emotion will have to be recognized in terms of continuous time, continuous valued dimensional affect in three dimensions: valence, arousal and dominance. Depression will have to be estimated in terms of a widely accepted clinical self-report questionnaire, the Beck Depression Inventory II.

Besides participation in the Challenge we are calling for papers addressing the overall topics of this workshop, in particular works that address the differences between audio and video processing of emotive data, and the issues concerning combined audio-visual emotion recognition.