Day 2 :
Keynote Forum
Vijayan K Asari
University of Dayton, USA
Keynote: A nonlinear manifold learning strategy for robust face recognition
Time : 10:05-10:35
Biography:
Dr Vijayan Asari is a Professor in Electrical and Computer Engineering and Ohio Research Scholars Endowed Chair in Wide Area Surveillance at the University of Dayton, Dayton, Ohio, USA. He is the director of the Center of Excellence for Computer Vision and Wide Area Surveillance Research (Vision Lab) at UD. As leaders in innovation and algorithm development, UD Vision Lab specializes in object detection, recognition and tracking in wide area surveillance imagery captured by visible, infrared, thermal, hyperspectral, LiDAR (Light Detection and Ranging) and EEG (electroencephalograph) sensors. Dr Asari's research activities include development of novel algorithms for human identification by face recognition, human action and activity recognition, brain signal analysis for emotion recognition and brain machine interface, 3D scene creation from 2D video streams, 3D scene change detection, and automatic visibility improvement of images captured in various weather conditions. Dr Asari received his BS in electronics and communication engineering from the University of Kerala, India, and M Tech and PhD degrees in Electrical Engineering from the Indian Institute of Technology, Madras. Prior to joining UD in February 2010, Dr Asari worked as Professor in Electrical and Computer Engineering at Old Dominion University, Norfolk, Virginia for 10 years. Dr Asari worked at National University of Singapore during 1996-98 and led a research team for the development of a vision-guided microrobotic endoscopy system. He also worked at Nanyang Technological University, Singapore during 1998-2000 and led the computer vision and image processing related research activities in the Center for High Performance Embedded Systems at NTU. Dr Asari holds three patents and has published more than 500 research papers, including 85 peer-reviewed journal papers in the areas of image processing, pattern recognition, machine learning and high performance embedded systems. Dr Asari has supervised 22 PhD dissertations and 35 MS theses during the last 15 years. Currently 18 graduate students are working with him in different sponsored research projects. He is participating in several federal and private funded research projects and he has so far managed around $15M research funding. Dr. Asari received several teaching, research, advising and technical leadership awards. He is a Senior Member of IEEE and SPIE, and member of the IEEE Computational Intelligence Society. Dr Asari is the co-organizer of several SPIE and IEEE conferences and workshops.
Abstract:
The human brain processes enormous volumes of high-dimensional data for everyday perception. To humans, a picture is worth a thousand words, but to a machine, it is just a seemingly random array of numbers. Although machines are very fast and efficient, they are vastly inferior to humans for everyday information processing. Algorithms that mimic the way the human brain computes and learns may be the solution. In this paper we present a theoretical model based on the observation that images of similar visual perceptions reside in a complex manifold in a low-dimensional image space. The perceived features are often highly structured and hidden in a complex set of relationships or high-dimensional abstractions. To model the pattern manifold, we present a novel learning algorithm using a recurrent neural network. The brain memorizes information using a dynamical system made of interconnected neurons. Retrieval of information is accomplished in an associative sense. It starts from an arbitrary state that might be an encoded representation of a visual image and converges to another state that is stable. The stable state is what the brain remembers. In designing a recurrent neural network, it is usually of prime importance to guarantee the convergence in the dynamics of the network. We propose to modify this picture: if the brain remembers by converging to the state representing familiar patterns, it should also diverge from such states when presented with an unknown encoded representation of a visual image belonging to a different category. That is, the identification of an instability mode is an indication that a presented pattern is far away from any stored pattern and therefore cannot be associated with current memories. These properties can be used to circumvent the plasticity-stability dilemma by using the fluctuating mode as an indicator to create new states. We capture this behavior using a novel neural architecture and learning algorithm, in which the system performs self-organization utilizing a stability mode and an instability mode for the dynamical system. Based on this observation we developed a self-organizing line attractor, which is capable of generating new lines in the feature space to learn unrecognized patterns. Experiments performed on various face lighting variant, pose variant and expression variant databases for face recognition have shown that the proposed nonlinear line attractor is able to successfully identify the individuals and it provided better recognition rate when compared to the state of the art face recognition techniques. These results show that the proposed model is able to create nonlinear manifolds in a multidimensional feature space to distinguish complex patterns.
Keynote Forum
Robert S Laramee
Swansea University, UK
Keynote: Visual analytics for big video data
Time : 10:35-11:05
Biography:
Robert S Laramee received a bachelors degree in physics, cum laude, from the University of Massachusetts, Amherst (ZooMass). He received a masters degree in computer science from the University of New Hampshire, Durham. He was awarded a PhD from Vienna University of Technology (Gruess Gott TUWien), Austria at the Institute of Computer Graphics and Algorithms in 2005. From 2001 to 2006 he was a researcher at the VRVis Research Center (www.vrvis.at) and a software engineer at AVL (www.avl.com) in the department of Advanced Simulation Technologies. Currently he is an Associate Professor in Data Visualizaton at Swansea University (Prifysgol Cymru Abertawe), Wales in the Department of Computer Science (Adran Gwyddor Cyfrifiadur). His research interests are in the areas of big data visualization, visual analytics, and human-computer interaction. He has published more than 100 peer-reviewed papers in scientific conferences and journals and served as Conference Chair of EuroVis 2014, the premiere conference on data visualization in Europe.
Abstract:
With advancements in multimedia and data storage technologies and the ever-decreasing costs of hardware, our ability to generate and store evermore video and other multimedia data is unprecedented. YouTube, for example, has over 1 billion users. However, a very large gap remains between our ability to generate and store large collections of complex, time-dependent video and multimedia data and our ability to derive useful information and knowledge from it. Viewing video and multimedia as a data source, visual analytics exploits our most powerful sense, vision, in order to derive information, knowledge and gain insight into big multimedia data sets that record complicated and often time-dependent events. This talk presents a case study of state-of-the art visualization and visual analytics techniques applied to video multimedia in order to explore, analyze, and present video data. In this case, we show how glyph-based visualization can be used to convey the most important information and events from videos of rugby games. The talk showcases some of visualizations strengths, weaknesses, and, goals. We describe inter-disciplinary case-study based on rugby sports analytics, where visual analytics and visualization is used to address fundamental questions-the answers of which we hope to discover in various large, complex, and time-dependent multimedia data