ERW 2015 - Special Issue - Emotion Recognition In The Wild

Springer Journal of Multimodal User Interfaces
---
Special Issue - Emotion Recognition In The Wild
---
The recent advancement of social media has given users a platform to socially engage and interact with a larger population. Millions of images and videos are being uploaded every day by users on the web from different events and social gatherings. There is an increasing interest in designing systems capable of understanding human manifestations of emotional attributes and affective displays. For inferring the affective state of the users in such images and videos, captured in real-world conditions, methods which can perform emotion analysis `in the wild' are required. Here, the term `in the wild’ signifies different environments/scenes and backgrounds noise, illumination conditions, head pose and occlusion. Automatic emotion recognition has made a significant progress in last two decades. However, such developed frameworks have been strictly employed to the data collected in controlled laboratory settings with frontal faces, perfect illumination and posed expressions. On contrary, the images and videos on the web have been captured in different, unconstrained environments and this poses a big challenge to automatic facial emotion recognition methods.
Recently, the First Emotion Recognition In The Wild (EmotiW) Grand Challenge and Workshop 2013 was organized, which brought together researchers working on emotion recognition. EmotiW2013, was successful and attracted 27 teams and including 9 final paper submissions. The event highlighted several challenges which need to be tackled for affect analysis in the wild. It is evident from the performance of the top performing teams at the EmotiW 2013 challenge that automated emotion recognition in the wild is a tough problem due to various limitations such as robust face detection and alignment, and environmental factors such as illumination, head pose and occlusion. Similarly, recognising vocal expression of affect in real-world conditions is equally challenging. Moreover, as the data (Acted Facial Expressions In The Wild database) has been captured from movies, there are many different scenes with very different environmental conditions in both audio and video, which will provide a challenging testbed for state-of-the-art algorithms, unlike the same scene/backgrounds in lab controlled data.
Therefore, it is worthwhile to investigate the applicability of multimodal systems for emotion recognition in the wild. There has been much research on audio only, video only and to some extent audio-video multimodal systems but for translating emotion recognition systems from laboratory environments to the real-world multimodal benchmarking standards are required. The Second Emotion Recognition In The Wild Challenge and Workshop is being organized this year. This special issue, will comprise of the invited best papers from EmotiW 2013 and 2014 plus papers from researchers describing methods addressing one or more of the challenges from the scope below:
Scope
- Multimodal emotion recognition in the wild
- Analysis paper based on EmotiW challenge data
- Vision based temporal emotion analysis in the wild
- Audio based emotion recognition
- New emotional data corpus representing real-world condition
- Emotion recognition applications in arts, entertainment and healthcare
Timeline
Paper submission deadline: 15thDecember 2014
Acceptance notifications: 1stMarch 2015
Final papers: 1stMay 2015
Submission Instructions
Submissions should be around 8-12 pages and must not have been previously published, with the exception that substantial extensions of the invited EmotiW 2013 and 2014 (at least 30% new content) can be considered.
Authors are requested to follow instructions for manuscript submission to the Journal of Multimodal User Interfaces (http://www.springer.com/computer/hci/journal/12193) and to submit manuscripts at the following link: http://www.editorialmanager.com/jmui/. The article type to be selected is “Special Issue S.I. :EmotionWild”.
Guest Editors
Abhinav Dhall, University of Canberra/Australian National University
abhinav.dhall-AT-anu.edu.au
Roland Goecke, University of Canberra/Australian National University
roland.goecke-AT-ieee.org
Tom Gedeon, Australian National University
tom.gedeon-AT-anu.edu.au
Nicu Sebe, University of Trento
sebe-AT-disi.unitn.it