International Workshop on Video Segmentation in Computer Vision
2nd of November 2014, Singapore
in conjunction with Asian Conference on Computer Vision
Aims and scope

Video segmentation is the basis for many applications: object tracking, object recognition, 3D reconstruction, robot navigation, activity recognition, and video retrieval. Despite many available video segmentation methods, the use of video segmentation as an early processing step in video analysis lags behind the use of image segmentation for image analysis. Most approaches are based on extending single image segmentation techniques to multiple frames, exploiting the fact that there is redundancy along the time axis and that the motion field is smooth. While this can be attempted by analyzing individual image frames independently, video provides rich additional cues beyond a single image. These cues include object motion, temporal continuity, and long-range temporal object interactions, etc. The purpose of this workshop is to bring together an examination of recent advances in video segmentation with emerging problem formulations motivated by Computer Vision applications.

The workshop will consist of a combination of invited talks, oral presentations, and panel discussion. Submissions are invited from all areas of computer vision relevant for video segmentation. Topics of interest include, but are not limited to:

  • Video segmentation
  • Hierarchical segmentation
  • Superpixel segmentation
  • Motion segmentation
  • Video cosegmentation
  • MRF for video segmentation
  • Semantic labeling in video segmentation
  • Evaluation metrics
  • Applications of video segmentation

All manuscripts will be subject to a double-blind review process. In the proceedings, accepted papers will be allocated 14 pages plus references.

Important dates
  • Submission site open: 15th of June 2014
  • Full paper submission (via ACCV main conf.): 20th of June 2014, 23:59 PDT (GMT-7)
  • Full paper submission (Workshop only): 10th of September 2014, 23:59 PDT (GMT-7)
  • Notification of acceptance: 25th of September 2014
  • Camera-ready paper: 1st of October 2014
  • Workshop: 2nd of November 2014
Keynote Speakers
  • Jianfei Cai, Nanyang Technological University
    Title: Geometry-aware Unsupervised / Semi-supervised Segmentation of Image Sequences
    Abstract: Video or image sequence segmentation has been studied for decades, but until today it is still regarded as a challenging problem, especially on achieving general object-level segmentation with temporal coherence and fast speed. In this talk, we present two of our recent works on segmenting image sequences, where there exist strong geometry constraints or geometry properties. In particular, we first consider the problem of precisely extracting the foreground object with silhouette coherence from a set of multi-view image sequence with a small amount of user interaction. Then, I will share with you our work on using shadow information for automatic real-time RGB-D video segmentation with some demo.

  • Joern Jachalsky, Technicolor
    Title: Superpixels for Video Segmentation
    Abstract: Superpixel algorithms represent a very useful and increasingly popular preprocessing step for a wide range of computer vision applications like segmentation, tracking, classification etc. Superpixel algorithms group spatially-coherent pixels that share similar features as e.g. color or texture into segments of rather homogeneous size and shape. This over-segmentation leads to a major reduction of image primitives, which results in an increased computational efficiency for subsequent processing steps and allows for more complex algorithms. This is especially interesting and beneficial for applications processing video content. In this talk, the basic ideas of superpixels are presented and an overview of approaches for still images as well as video content is given marking important steps in the research on superpixels. Thereby, the challenges for superpixels for video content are discussed. Finally, it is highlighted how superpixels for video content can be used for video segmentation with a special emphasis on interactive video segmentation.

  • Junsong Yuan, Nanyang Technological University