Author
Andrew B. Whinston
Bio: Andrew B. Whinston is an academic researcher. The author has contributed to research in topics: Officer. The author has an hindex of 2, co-authored 2 publications receiving 32 citations.
Topics: Officer
Papers
More filters
01 Jan 2009
5 citations
Cited by
More filters
01 Jan 2008
TL;DR: This special issue, which focuses on event analysis in broad problem domains, has witnessed the effectiveness of using both static and temporal information in event recognition from other video sources.
Abstract: Event analysis in videos is a critical task in many applications. Activity recognition that aims to recognize actions from video and in particular abnormal event recognition in surveillance video has received significant attention from the research community. In this special issue, we focus on event analysis in broad problem domains. Event recognition in specific domains, such as highlight detection in sports videos, has attracted much interest in the past decade. Recently, due to the emergence of online video search, the research community has become interested in event content analysis for both broadcast and user-generated videos. For news videos, Large-Scale Concept Ontology for Multimedia (LSCOM) has defined 56 event/activity concepts, covering a broad range of events such as airplane flying, car crash, riot, people marching, and so on. Researchers have also started to investigate event recognition from other video sources, such as education videos and medical videos. For these applications, we have witnessed the effectiveness of using both static and temporal information.
428 citations
01 Jan 2011
TL;DR: In the Special Issue on Multifunctional Circuits and Systems for Future Generations of Wireless Communications, the search is looking for circuits and systems solutions for multiple communication standards.
Abstract: The explosive demand in wireless-capable devices, especially with the proliferation of multiple standards, indicates a great opportunity for adoption of wireless technology at a mass-market level. The communication devices of both today and the future will have not only to allow for a variety of applications, supporting the transfer of characters, audio, graphics, and video data, but they will also have to maintain connection with many other devices rather than with a single base station, in a variety of environments. Moreover, to provide various services from different wireless communication standards with higher capacities and higher data-rates, fully integrated and multifunctional wireless devices will be required. Multifunctional circuits and systems can be made profitable by a large scale of integration, elimination of external components, reduction of silicon area, and extensive reuse of resources. Integration of (Bi)CMOS transceiver RF front-end and analog baseband circuits with computing CMOS circuits on the same silicon chip further reduces costs of multifunctional mobile devices. However, as batteries continue to determine the lifetime and size of mobile equipment, further extension of capabilities of wearable and wireless devices will depend critically on the integrated circuits and systems solutions. The demand for multifunctional and multi-mode wireless-capable devices is accompanied by many significant challenges at system, circuit, and technology levels. In the Special Issue on Multifunctional Circuits and Systems for Future Generations of Wireless Communications, we are looking for circuits and systems solutions for multiple communication standards. Examples of topics qualifying for the special issue include: • Adaptive radio circuits and systems • Multifunctional multistandard multi-band circuits and systems • Software-defined radio circuits and systems • Cognitive radio circuits and systems • Low-voltage low-power RF and analog circuits for future generations wireless systems • Ultra Wide Band circuits and systems
133 citations
••
01 Jan 2009
TL;DR: A methodology to exploit a specific type of domain knowledge in order to find tighter error bounds on the performance of classification via Support Vector Machines by using the ellipsoid method from optimization literature.
Abstract: In this study we describe a methodology to exploit a specific type of domain knowledge in order to find tighter error bounds on the performance of classification via Support Vector Machines. The domain knowledge we consider is that the input space lies inside of a specified convex polytope. First, we consider prior knowledge about the domain by incorporating upper and lower bounds of attributes. We then consider a more general framework that allows us to encode prior knowledge in the form of linear constraints formed by attributes. By using the ellipsoid method from optimization literature, we show that, this can be exploited to upper bound the radius of the hyper-sphere that contains the input space, and enables us to tighten generalization error bounds. We provide a comparative numerical analysis and show the effectiveness of our approach.
19 citations