All areas that enable, optimise or enhance Speech in mobile and pervasive environments and devices. Possible areas include, but are not restricted to:
Robust Speech Recognition in Noisy and Resource-constrained Environments
Memory/Energy Efficient Algorithms
Multimodal User Interfaces for Mobile Devices
Protocols and Standards for Speech Applications
Distributed Speech Processing
Mobile Application Adaptation and Learning
Prototypical System Architectures
User Modelling
Submissions
We invite position papers (upto 8 pages - shorter papers are also welcome). Electronic submission is required. Submissions should be formatted according to ACM SIG style. All submissions should be in PDF (preferred) or Postscript format. If any of these requirements is a problem for you, please feel free to contact the workshop organisers.
We also welcome participation without paper submission.
Position papers must be submitted via the conference submission web site,
http://www.easychair.org/SiMPE07.
For any comments regarding submissions and participation, contact:
namit AT in DOT ibm DOT com, rnitendra AT in DOT ibm DOT com
Note: Registration for this workshop is included in the MobileHCI'07 conference registration fee. Alternatively, participants can register only for the workshop - this has a $0 fee.
Seed Questions
How to do speech recognition in noisy environments ?
How to make voice UIs flexible and adaptive ?
Are there any novel and easier ways to handle multiple
languages and dialects ?
How do we construct speech systems with small footprints
of memory and power consumption ?
How can we distribute processing more efficiently given
the increased available computing power on handhelds ?
How do we trade this off with a remote server to
conserve energy ?
Do we need shorthand languages for the mobile speech
UI ?
How do we make such devices adapt automatically to
the user, task and environment ?
What novel applications and services can be deployed
on such devices ?
What are the right learning models for such settings ?
What does the architecture of future systems looks like
?
How can we design scalable speech applications ?
How can we leverage context (such as location) to
make more intelligent UIs that reduce the cognitive
burden of semi-literate/illiterate users ?
What are the acoustic implications ? The major challenge
in deploying ubiquitous ASR is that the operating environments
may change rapidly leaving the ASR
system very vulnerable. To increase ASR system robustness
in varying environmental noise context, we
discuss schemes that are applicable to a specific environment context.
