Scenery of Shanghai

Person Search and Verification from Rich Media Data

With the explosive dissemination of diverse video capturing devices and mobile networks, we have entered an era where the staggering amount of multimedia data has far surpassed human ability to process all of it manually.

This Special Session will focus on new algorithms and systems that aim at finding persons according to their appearances, identities, attributes, and activities in such media streams. We encourage submissions that surpass traditional person re-identification systems and tackle the new challenge (or opportunity) presented by the availability of rich multimodal data in this space by using: novel representations, learning paradigms, multi-modal frameworks, algorithms designed to work at scale.

We also welcome novel collections that could be used to further spur research and measure performance of intelligent systems for person search and verification with different application backgrounds, e.g., intelligent surveillance, image/video content-based retrieval, social relation mining, cross-media retrieval.

Scope and Topics

  • Person verification from image and video
    • New models (e.g., feature representations and/or distance metrics) for capturing people appearance variations.
    • New algorithms of inference and learning for matching persons under different scenes.
    • New applications centering at person search and verification, e.g.
      • Mobile-based persons of interest retrieval
      • Person attributes and semantics understanding
      • Person re-identification in surveillance
      • Face identity recognition
  • Person search across diverse medias
    • New approaches of multi-modal person search
    • New deep learning frameworks to fuse heterogeneous data for person search across diverse medias
    • New approaches for cross-media news retrieval
    • New benchmarks and applications for person search with rich media data
  • Industry demos and new applications
    • ew systems or live demos in the related topics.
    • New applications of person search or verification


Liang Lin, Sun Yat-Sen University, China

Michele Merler, IBM T. J. Watson Research Center

Liangliang Cao, IBM T. J. Watson Research Center


Deadline: January 25, 2015, 11:59 PM PST

Please submit your work using the EasyChair conference website.

Sponsors and Partners

Instant Cash Loans Online Near Me No Credit Check