Quality Assessment and Enhancement for Multimedia Visual Signals

Nowadays, multimedia visual signals, including 2D, 3D/stereopsis, virtual reality (VR), augmented reality (AR), light field, and point clouds, have become unprecedentedly popular. It has witnessed a wide range of potential applications in television, automatic driving, medical diagnosis, cultural relic protection, business marketing, transportation, gaming, and education, among many others. The development of multimedia technologies including acquisition, compression, transmission, reconstruction, and display, has made great progress, providing new ways for manufacturers and consumers to generate, use and interact with massive visual information. However, there exist many quality problems and challenges in multimedia signal processing. In this special session, the goal is to publish original research papers related to the theories, models, and algorithms about the quality assessment and enhancement for multimedia visual signals with widespread applications.
The main topics of interest include, but are not limited to:

  • Subjective quality assessment methodologies
  • Objective quality assessment models
  • Physiological experiments
  • Relationship between human behavior and perceptual quality
  • New synthetic or real multimedia quality databases
  • Visual quality enhancement, such as super-resolution, denoising, restoration
  • A survey of multimedia visual quality assessment and enhancement technology
  • Perception-driven visual signal processing methods
  • Visual attention modeling and its applications in perceptual signal processing
  • Dr. Wei Zhou (wei.zhou@uwaterloo.ca): Department of Electrical and Computer Engineering, University of Waterloo, Canada
  • Dr. Guanghui Yue (yueguanghui@szu.edu.cn): School of Biomedical Engineering, Shenzhen University, China
  • Dr. Xiongkuo Min (minxiongkuo@sjtu.edu.cn: Department of Electronic Engineering, Shanghai Jiao Tong University, China
  • Dr. Jesús Gutiérrez (jesus.gutierrez@upm.es): Image Processing Group, Universidad Politécnica de Madrid, Spain

Emerging Computational Imaging and Multimedia Signal Processing Techniques for the Metaverse

The world is witnessing a dramatic change in the way that people experience it due to the rise of Metaverse. Social life and economic activities are redefined through the blurred boundary of digital and physical worlds. To pursue a realistic immersive experience, a range of cutting-edge technologies are incorporated to empower Metaverse, such as artificial intelligence (AI), extended reality (XR), 3D modeling and reconstruction, etc. These related technologies are all data-driven. Thus, the quality of data acquisition (imaging) and the richness of data modality essentially determine the Metaverse’s development.

Recent advances in computational imaging enable to extend the data acquisition capability and improve imaging quality with fewer hardware resources. Some examples include controlling depth of field, extending field of view, noise removal, super-resolution, etc. Besides, multiple immersive media formats can provide a more comprehensive and informative description of scenes, achieving immersive experience. Typical formats include light field, 360 video, point cloud, multiview, multi-focus, structured light, etc. This special session is devoted to the publications of high-quality papers on technical developments and practical applications around “Computational Imaging and Multimedia Signal Processing for the Metaverse”. In this session, we would like to aim at connecting a broad and cross- disciplinary research, including computational optical imaging, immersive media, cross-modal multimedia fusion, signal processing, and other fields.

Topics of interest include but are not limited to:

  • Computational optical sensing and imaging and its applications in Metaverse
  • Data acquisition and processing of immersive media formats to achieve 6DoF, e.g., light field, 360video, multi-focus, and point cloud
  • Neural-based models for immersive media format processing, e.g., NeRF based refocusing,rendering and compression, quality assessment and streaming
  • Cross-modal multimedia signal processing, e.g., Diffusion model-based text-to-3D, depth-RGB-spectral data fusion for object detection and action recognition
  • Efficient multimedia communications (e.g., low-latency, ultra-low bitrate compression)
  • Dr. You Yang (yangyou@hust.edu.cn): School of Electronic Information and Communications, Huazhong University of Science and Technology, Wuhan, China
  • Dr. Kejun Wu (kejun.wu@ntu.edu.sg): School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore
  • Dr. Giuseppe Valenzise (giuseppe.valenzise@l2s.centralesupelec.fr): University Paris-Saclay, L2S, CNRS, CentraleSupelec, Paris, France

Recent advances in movement studies

Motion measurement signal analysis is useful for studying living being’s behavior, 3D motion measurements, optic flow, object detection, tracking, people and scene recognition or calibrating devices. In the health sciences, motion analysis can be used to detect deficiencies in the sensorimotor system at an early stage, to prevent diseases or risk behaviors related to aging. Crowd modeling is useful to reproduce and better understand complex movement phenomena. In this context, there are many sensors for the study of motion and data fusion is then essential to better understand the acquired signals and to analyze motion in all its complexity. On the one hand, signal processing methods have proven to be efficient for problems involving limited data volumes, on the other hand, deep learning methods have achieved remarkable performances for problems with large data volumes while allowing to analyze the event in real time.

This special session aims to follow up on the latest developments, research findings and trends on the following topics, including but not limited to:

  • Human detection and tracking
  • Human movement
  • Motion measurement
  • Action quality evaluation
  • Interaction recognition
  • Action recognition
  • Pose estimation
  • Scene recognition
  • Simultaneous Localization and Mapping (SLAM)
  • Object detection
  • Object tracking
  • Temporal action localization
  • Organism tracking and movement analysis
  • Optical flow
  • Dr. Baptiste Magnier (baptiste.magnier@mines-ales.fr): Mines-Telecom Institute Alès, Alès, France