ARIA - Quality-Adaptive Media-Flow Architectures for Sensor Data Management
  NSF Award : IIS-0308268
Home People Research Publications Directions Screenshots
     

The inter-disciplinary team brings together experts in the areas of multimedia databases and information fusion (Candan), hardware/software systems and optimization (Chatha), and media processing and adaptive feature extraction (Sundaram). The expertise and experience of the researchers as relates to this project are detailed below.


K. Selcuk Candan, Assoc. Prof., Computer Science and Engineering

Candan has worked extensively on the integration and representation of heterogeneous information and information sources. In recent years, he has been involved in various multimedia information management projects, including CHIMP (Collaborative Heterogeneous Interactive Multimedia Presentations), SEMCOG (Semantics and Cognition Based Media Retrieval), WebDB, and HUMOR (HeterogeneoUs Multimedia Object Replication). His recent research includes development of media view management and dynamic caching, similarity measures, data structures, and algorithms for efficient storage and retrieval of imprecise information. He has also developed quality based object fusion and routing algorithms. He is responsible for the investigations and design of the data management, fusion, and optimization techniques in this project. He brings to the team his expertise in multimedia systems and data management. More at http://www.public.asu.edu/~candan

 


Karamvir S. Chatha, Asst. Prof., Computer Science and Engineering

Chatha has a strong background in system-level hardware/software co-design and optimization. His Phd research at the University of Cincinnati concentrated on the development a system-level design environment called STELLAR for specification and timing optimization of multimedia applications. As part of his Phd research, he developed a C++ class library called NOVA for specification of application functionality, target architecture and performance constraints. He also developed a number of techniques for performance evaluation, execution time minimization and throughput maximization of multimedia applications that are mapped to heterogeneous architectures. His paper titled "Hardware-Software Codesign for Dynamically Reconfigurable Architectures" was recognized by the Best Paper Award at the Field Programmable Logic (FPL) and ApplicationsWorkshop, 1999. More at http://www.eas.asu.edu/~kchatha

 


Hari Sundaram, Asst. Prof., Computer Science and Engineering

Sundaram has worked extensively on the multimedia content analysis problems - representation of audio visual data, structure detection and summarization, as well problems relating to multimedia retrieval [116]. His recent projects include VideoQ, a novel content-based video retrieval framework, and KIA, a multimedia content analysis project. KIA dealt with segmentation problems (new signal processing algorithms for audio and visual segmentation), generative models structure detection, and a entity-utility based constrained minimization framework for multimedia summarization. He has received a best paper award for his work on VideoQ, in IEEE Transactions on Circuits and Systems for Video technology and a best paper nomination in ACM SIG Multimedia 2002. He was also awarded the Eliahu I Jury award for best dissertation at Columbia University. He brings to the group, his expertise in multimedia analysis. More at http://ame2.asu.edu/faculty/hs

 

   


Lina Peng, Ph.D Candidate, Computer Science and Engineering

Lina Peng is a doctoral candidate in Computer Science at Arizona State University. She received her Master degree from University of Washington, Seattle, in 2002, her first Master degree and Bachelor degree from Beijing University of Aero. and Astro., Beijing, China, in 2000 and 1997. Lina was awarded Louis and Katherine Marsh Memorial Fellowship, Guanghua graduate scholarship, and outstanding undergraduate student scholarship four times in a row. Lina's research interests include developing quality-adaptive architectures to manage media data, quality-based object routing algorithms in media processing workflows, generalized models to assess object quality, and quality-based object fusion algorithms. She has published several conference and journal papers on the related topics. With her demonstrated performances in education and scientific publication, she contributes to the modeling of workflows with adaptive QoS assurance and object quality assessments, and the specification and visualization of media processing workflows.

 

 
   
Gisik Kwon, Ph.D Candidate, Computer Science and Engineering

Gisik's research interests include the efficient and effective distributed computing and networking: in particular, how to take advantage of the resource heterogeneity inevitable in a large distributed network like Internet; how to adapt the unpredictable changes in the network which assures the run-time QoS specified in the service requirement. The projects he has been working in ASU cover the state-of-the-art distributed networking and computing technologies to solve such fore-mentioned adversities, resulting in the peer-to-peer file sharing systems, PASS(Peer-to-peer Asymmetric information Sharing System) and BYPASS(the overlay version of PASS), and the decentralized workflow processing systems (DANS, extended from ARIA workflow processing system). With empirical network programming skills and academic backgrounds thereof, he makes contributions to design and implement ARIA kernel, a multi-threaded network program supporting the long-term workflow execution along with the adaptive resource utilization and QoS enforcement.

 

 
    Affiliated Researchers  

Maria Luisa Sapino

Since mid-90s, Prof. Sapino (Associate Professor, U. of Torino) has been applying Logic Programming and AI techniques to the challenges associated with database access control, and with heterogeneous and multimedia data management. In particular, she developed novel techniques and algorithms for similarity based information retrieval, content based image retrieval, web accessibility for users who are visually impaired. She also focused on temporal and synchronization aspects of distributed multimedia presentations in the presence of resource constraints. Her current focus includes the modeling and investigation of various aspects of ambient intelligence systems. More at http://www.di.unito.it/~mlsapino/


Kyung D. Ryu

Dr. Ryu (Research Staff Member, IBM TJ Watson Research Center) has been working on resource scheduling and dynamic adaptation in distributed and parallel computing systems. His work in the Active Harmony developed fine-grain cycle stealing, a novel technique for utilizing by utilizing idle processor cycles from non-dedicated networks of workstations. For this project, he also developed the a set of scheduling mechanisms for resource restrictions in operating systems. He is currently working on serverless peer-to-peer distributed computing. The related projects includes COIN (Consumer-Oriented computing Infrastructure) and PASS (Peer-to-peer Asymmetric information Sharing System). His expertise in runtime adaptation in distributed computing contributes to runtime refining issues of ARIA media flow planning.
 




1st Intl. Workshop on Ambient Intelligence, Media, and Sensing
2007
Istanbul, Turkey