This poster proposes two research topics. The first topic deals with visual analytics for multimedia quality evaluation. It tries to relate subjective and objective evaluation of multimedia content quality. It also tries to answer the question on what are the new approaches for visual quality modeling under uncertainty. The second topic deals with automated visual analytics for security applications. It combines visual analytics made by humans and computers. It also tries to answer the question on how to apply visual and video analytics for security/safety applications and how take decisions when supplementary information is a scarce and uncertain resource.
The communication presentations (keynote talks, facilitated break-out sessions, posters) will be continuously in May and June.
|CHIST-ERA Conference 2016 - M. Leszczuk.pdf||913.04 KB|
The LNE is engaged in independant system evaluation of information processing systems, in embedded contexts or not. We propose systemic evaluations with process decomposition, going from physical sensors to processing. We also try to develop methodologies for data collection and curation for system evaluation and experiment reproducibility.
Dr. Sophie Rosset is a CNRS Senior Researcher (DR2) at LIMSI (Orsay, France), where she is responsible for the research activities "Robust analysis of spoken language and dialog systems" in the Spoken Language Processing group. Her graduate and post-graduate research focus on spoken language dialog systems, interactive spoken question-answering and information extraction and management from large-scale textual and noisy data. She is the author or co-authopr of original methods and tools to manage dialog and extract information in an error-prone environment. She has also co-designed methods to acquire, develop, annotate and validate corpora for various natural language processing. In the past few years, she has been prime contributor to the LIMSI participation in Question Answering on Speech Transcript evaluations (QA@CLEF) where the LIMSI systems obtained the best results in 2008 and 2009 (French, English, Spanish), she led the first project on dialog system in open-domain for which she designed original methods to represent semantic and technical information conveyed by the user-system interaction. She was responsible of the Named Entity activities within the Quaero program (on both technological and resources creation sides) and the French Edylex project. She is responsible for the Conversational Agent workpackage within the Patient Genesys project. She is the author or co-author of over 100 peer-reviewed publications.
Incremental learning and specifically interactive and incremental learning will lead to systems able to learn through natural language interaction and examples. This will allow a deep true adaptation to users or to be responsive to specific needs. After a brief overview of machine learning and natural language processing intertwining with incremental learning, I will present an experiment on interactive and incremental learning based on formal analogy reasoning. In this experiment we were interested in an intelligent assistant that can learn from the user a new task or a new domain. The evaluation procedure will be presented.
|CHIST-ERA Conference 2016 - Sophie Rosset.pdf||166.28 KB|
Existing systems are stubbornly limited to do what they are initially programmed for, unable to adapt to newly arising user needs. Learning is done often in an off-line and supervised manner. Adaptation remains costly and time-consuming and thus a major bottleneck to meet the individual needs of the widest possible range of users. It is time to propose new methods to develop devices able to learn through natural communication with humans, leading to the dreamt goal of an intelligent, communicative and then evolving machine which is able to adapt itself to the user's changing needs and practices.
Torsten Möller is a professor at the University of Vienna, Austria, since 2013. Between 1999 and 2012 he served as a Computing Science faculty member at Simon Fraser University, Canada. He received his PhD in Computer and Information Science from Ohio State University in 1999 and a Vordiplom (BSc) in mathematical computer science from Humboldt University of Berlin, Germany. He is a senior member of IEEE and ACM, and a member of Eurographics. His research interests include algorithms and tools for analyzing and displaying data with principles rooted in computer graphics, human-computer interaction, image processing, machine learning and visualization.
He heads the research group on Visualization and Data Analysis. He served as the appointed Vice Chair for Publications of the IEEE Visualization and Graphics Technical Committee (VGTC) between 2003 and 2012. He has served on a number of program committees and has been papers co-chair for IEEE Visualization, EuroVis, Graphics Interface, and the Workshop on Volume Graphics as well as the Visualization track of the 2007 International Symposium on Visual Computing. He has also co-organized the 2004 Workshop on Mathematical Foundations of Scientific Visualization, Computer Graphics, and Massive Data Exploration as well as the 2010 Workshop on Sampling and Reconstruction: Applications and Advances at the Banff International Research Station, Canada. He is a co-founding chair of the Symposium on Biological Data Visualization (BioVis). In 2010, he was the recipient of the NSERC DAS award. He received best paper awards from IEEE Conference on Visualization (1997), Symposium on Geometry Processing (2008), and EuroVis (2010), as well as two second best paper awards from EuroVis (2009, 2012). In 2016 he received the Teaching Award from the University of Vienna.
Modern science is driven by computers (computational science) and data (data-driven science). While visual analysis has always been an integral part of science, in the context of computational science and data-driven science it has gained new importance. In this talk I will demonstrate novel approaches in visualization to support the process of modeling and simulations. Especially, I will report on some of the latest approaches and challenges in modeling and reasoning with uncertainty.
Visual tools for ensemble analysis, sensitivity analysis, and the cognitive challenges during decision making build the basis of an emerging field of visual data science which is becoming an essential ingredient of computational thinking.
In any intelligent systems, knowledge is kept evolving, no matter this knowledge comes from domain experts or is discovered from machine leaning/data mining approaches. A fundamental question is how to manage and validate such evolving knowledge, what is the relationship between existing and newly acquired knowledge. We have developed knowledge fusion/updating and validation algorithms that precisely aim to meet these challenges. Knowledge fusion allows us to merge knowledge from different sources whilst knowledge updating enables us to revise existing knowledge in light of new knowledge. We have also developed computational viable software to identify inconsistencies among multiple knowledge bases to validate knowledge.
- ‹ previous
- 2 of 2