You are here

CVIU Special Issue on Image and Video Retrieval Evaluation

CALL FOR PAPERS

Computer Vision and Image Understanding

Special Issue on Image and Video Retrieval Evaluation

Guest Editors:

  • Allan Hanbury, Vienna University of Technology, Austria
  • Henning Müller, University and Hospitals of Geneva, Switzerland
  • Paul Clough, The University of Sheffield, UK

Important Dates:

Manuscript submission: 30 April 2008 (extended deadline)
Acceptance Notification: 1 September 2008
Final Manuscript Due: 1 November 2008
Special Issue to Publisher: 15 January 2009
Expected Publication Date: Mid 2009

Summary:

Visual information is ubiquitous and the amount produced with cheap digital cameras is rising strongly. To better manage this information, content-based image and video retrieval has been proposed for general retrieval as well as in specialised domains. While many techniques have been developed for image and video retrieval, one of the problems is that many published approaches are very difficult to compare to each other as varying databases, performance measures, and methodologies are used.

In order to mitigate this problem, there has been increased focus on the evaluation of visual information retrieval systems in the last years. Similar to the text retrieval domain many years ago, several visual information retrieval benchmarks with a varying focus have been created and run. This evaluation is particularly important as many visual information retrieval techniques and systems are at the point where they are ready to leave the academic field and become integrated into commercial prototypes and products. This requires techniques not only to be interesting as theoretical approaches but also to be comparable with respect to performance obtained.

The role of this special issue is to fill the need for a comprehensive overview of the current visual information retrieval evaluation activities. This is envisioned to include evaluation campaigns as well as activities at a lower level, such as dataset creation, innovative approaches to ground truth collection and relevance judgements, discussion of evaluation metrics and the planning of “realistic” queries and user models.

Scope:

The scope of this special issue is to cover all aspects of the evaluation of visual information retrieval.

Topics of interest include (but are not limited to):

  • Evaluation campaigns
  • Evaluation techniques
  • Evaluation metrics
  • Datasets and dataset creation
  • Query topic creation
  • Ground truth collection
  • Relevance judgements
  • Evaluation of image and video features for retrieval
  • Challenges in image and video retrieval evaluation
  • User requirement modelling and user modelling
  • Evaluation of user interfaces to search engines

Submission Procedure:

Papers should be appropriate for journal publication. Submissions should follow the guidelines set out by the CVIU.

When submitting papers, the authors should select the specific article type indicated AND include a line in their cover letter as a failsafe [example: "This manuscript is submitted to the Special Issue on Image and Video Retrieval Evaluation"].

All papers should be submitted via the CVIU web-site with Article Type ‘Special Issue: Image and Video Retrieval’
http://ees.elsevier.com/cviu/

Full author guides (Preparation of Manuscript, Copyright and Permissions, Author Inquiries, Submission of Manuscripts) and on-line submission links can be found at the above link.

All papers will be peer reviewed following the CVIU reviewing procedures.

Contacts:

Please address all correspondence concerning this special issue to the Guest Editors:
Allan Hanbury (hanbury@prip.tuwien.ac.at),
Henning Müller (henning.mueller@hevs.ch),
Paul Clough (p.d.clough@sheffield.ac.uk)