Aesthetic and Emotions in Multimedia

Subjective notions such as aesthetics and emotions are central to human-centric perception. Even with the use of multimedia technology, such notions remain realistic and plausible. Unlike objective and tangible objects, these concepts can be perceived differently from one person to another. With the recent advances in deep learning and computational models, it is inevitable that the analysis and understanding of subjective aspects such as image quality, aesthetics, and emotions from multimedia signals stand to benefit as well. It is also common for these aspects to be studied from various domains – from artistic media content (artworks, photographs, movies) to social media content (short clips, edited images) and advertisements, and how they are applicable to a range of applications such as psychology, sports, and surveillance. More recently, there is a strong interest in generative art and synthesis of multimedia content from alternative inputs such as text – how we evaluate the quality, aesthetics and emotions invoked through generated media presents interesting opportunities for research.

Topics of interest of this special session include, but not limited to:

  • Analysis and assessment of aesthetics in multimedia data, including associated concepts such as interestingness, popularity, and viralness
  • Analysis and prediction of emotions in multimedia data
  • Deep learning techniques for representation of subjective concepts of aesthetics and/or emotions
  • Multimedia content retrieval driven by aesthetic and/or emotions
  • Style transfer and personalized enhancement of multimedia content driven by aesthetic and/or emotions
  • Synthesis and generation of multimedia content driven by aesthetic and/or emotions
  • Subjective evaluation techniques and other human-in-the-loop methodologies
  • Applications and new datasets relating to aesthetics and emotions in multimedia

For authors:

1) Details at Call for Special Sessions
2) Paper templates and paper submission system at Instructions for Authors


John See

Lai-Kuan Wong

Wei-Ta Chu

Giuseppe Valenzise