Machine Vision and Applications

Special issue on:

Multimedia Event Detection


The amount of multimedia data being recorded, stored and shared is increasing daily, and the total sum of data is astronomical. For example,  just on the site YouTube, video data is being uploaded at the rate of one hour a second or 31 million hours a year*. This yields practical problems when analysts or laypersons want to extract information from this huge data collection. What is needed are tools for automatic tagging, indexing and searching of multimedia data. This is clearly a grand challenge and needs to be divided into manageable sub-problems. One of the interesting sub-problems is automatic event detection in videos. Recently some success has been reported for event detection in structured surveillance scenarios using computer vision. But a far bigger challenge is to generalize event detection to unconstrained settings. To achieve this end, it is widely believed that exploiting multiple modalities (imagery, audio, text, speech, etc.) jointly is more likely to be successful as opposed to using a single modality. The aim of this special issue is to promote research on the important topic of multimedia event detection where image/video data plays a central role.

Contributions related to all aspects of multimedia event detection are invited, including (but not limited to):

-       Representation, modeling and detection of multimedia events

-       Exploitation of existing/prior knowledge and knowledge-bases for event detection

-       Information fusion from audio, visual and text modalities

-       Multimedia event search tools and strategies

-       Human-in-the-loop: semi-automatic multimedia event detection

-       Indexing techniques for fast event detection

-       Metrics for probing multimedia events

-       Definition and extraction of generic meta data

-       Novel video-based event detection

-       Interactive tools for navigating multimedia events

-       Benchmark dataset construction for multimedia event detection




Guest Editors:

Manuscript Due:

First notification:

Final Decision:

Publication Date:

Oct. 20, 2012

Dec. 20, 2012

Apr. 15, 2013

Sep. 1, 2013

Thomas B. Moeslund, Aalborg University, Denmark

Omar Javed,  SRI International, Sarnoff Corporation, USA

Yu-Gang Jiang, Fudan University, China

R. Manmatha, University of  Massachusetts, Amherst, USA




Submission and Selection:

Submissions should follow the official guidelines of Machine Vision and Applications, which can be found at Prospective authors should submit quality and original manuscripts that have not appeared, nor are under consideration, in any other journals. Manuscripts should be submitted electronically through MVA submission system at All papers will be reviewed by at least three expert reviewers in relevant fields. Decision will be made based on the novel scientific and technical contribution and their suitability to the scope of this MVA special issue.