Audio and Multimedia Methods for Large‐Scale Video Analysis
http://amva2012.icsi.berkeley.edu

First ACM International Workshop at ACM Multimedia 2012
29 October ‐ 2 November in Nara, Japan

***Extended submission deadline: July 15th  2012 ***

Media  sharing sites on the Internet and the one‐click upload ca‐
pability of smartphones have led to a deluge of online multimedia
content.  Everyday, thousands of videos are uploaded into the web
creating an ever‐growing demand for methods to make  them  easier
to  retrieve,  search,  and  index. While visual information is a
very important part of a video, acoustic information  often  com‐
plements  it.  This  is  especially true for the analysis of con‐
sumer‐produced, unconstrained videos from social media  networks,
such as YouTube uploads or Flickr content.

The diversity in content, recording equipment, environment, qual‐
ity, etc. poses significant challenges to the  current  state  of
the  art in multimedia analytics. The fact that this data is from
non‐professional and consumer sources means  that  it  often  has
little or no manual labeling. Large‐scale multi‐modal analysis of
audio‐visual material can help overcome this problem, and provide
training  and testing material across modalities for language un‐
derstanding, human action recognition, and  scene  identification
algorithms,  with  applications  in robotics, interactive agents,
etc. Speech and audio provide a natural modality to summarize and
interact  with the content of videos. Therefore, speech and audio
processing is critical for multimedia analysis that  goes  beyond
traditional classification and retrieval applications.

The  goal of the 1st ACM International Workshop on Audio and Mul‐
timedia Methods for Large‐Scale Video Analysis (AMVA) is to bring
together  researchers  and  practitioners  in this newly emerging
field, and to foster discussion on future directions of the topic
by providing a forum for focused exchanges on new ideas, develop‐
ments, and results. The aim is to build a strong community and  a
venue that at some point can become its own conference.

Topics include novel acoustic and multimedia methods for
  * video retrieval, search, and organization
  * video navigation and interactive services
  * information extraction and summarization
  * combination, fusion, and integration of the audio,
    visual, and other streams
  * feature extraction and machine learning on 'wild' data

Submissions: Workshop submissions of 4‐6 pages should be  format‐
ted  according to the ACM Multimedia author kit. Submission  sys-
tem link: https://cmt.research.microsoft.com/ACMMMWS2012

Important dates:
Workshop paper submission: July 1st, 2012 
Notification of acceptance: August 7th, 2012
Camera ready submission to Sheridan: August 15, 2012

Organizers:
Gerald Friedland, ICSI Berkeley (USA)
Daniel P. W. Ellis, Columbia University (USA)
Florian  Metze,  Carnegie‐Mellon  University (USA)

Panel Chair:
Ajay Divakarian, SRI/Sarnoff (USA)