MULTIMEDIA COMMUNICATIONS RESEARCH ACTIVITY 2003 The Multimedia Communications Department is actively involved in a number of research projects that study fundamental technologies to improve the processing and usage of multimedia information. These research projects are organized along the following themes: Indexing of stored multimedia information, which aims at improving the search and retrieval of multimedia information, and multimedia filtering, which can be used to automatically direct multimedia information to the right users ; Virtual reality imaging, which aims at creating virtual meeting rooms, and watermarking of images and video, which can ensure effective copyrighting of image and videos ;
Multimedia Information Processing Principal Investigator: Prof. Bernard MERIALDO • Highlights In 2003, the group has continued its participation to the two european projects, Spation and GMF4iTV, and and initiated a collaboration among French laboratories for a common participation to the TRECVideo experiments. • Research Themes The research activity within the Multimedia Information Processing group is organized along two themes: • Multimedia Indexing: We are studying mechanisms to describe the content of multimedia information, with a special focus on the processing of audio-video sequences. These mechanisms span a whole range of domains, from content extraction based on recognition techniques, content understanding based on parsing and multimedia content description. Our focus is on developing models for the description of multimedia content. For example, we are working on the automatic parsing of TV news recordings, and on the automatic construction of video summaries. • Information Retrieval and Filtering: We are studying mechanisms to automatically direct multimedia information to the right users. This includes several problems, from the description of the multimedia information content, the definition of user profiles, and the matching between document content and users. In particular, we are studying collaborative filtering techniques, which predict the importance of an information for a specific user based on previous evaluations of other users. An application of this technique is the Active Web Museum , which personalizes the visit of a user based on his personal preferences, and the opinion of previous visitors. • Multimedia Indexing • Video Summarization We are investigating techniques to build automatically video summaries. A video summary is a sequence of short but “important” clips that have been extracted from the original video. The essence of this problem is to determine what is important and to find where the important clips are. We have developed an original approach based on the concept of “Recollection” to define the quality of the summary as the performance of a simulated user on an realistic retrieval task. This allows us to define optimal summaries as those which maximize the user performance. The quality of the optimal summaries can then be measured with a realistic figure, rather than an abstract percentage. Our first approach were based on video only. In 2003, we have extensively studied several approaches for the combination of video and text. In particular, we have introduced constraints to optimize the information content visible on a given amount of display space. We have also started some experimentations on the combination of video and audio. This work is part of the European project SPATION, which aims a developing techniques and applications dedicated to the home network. The prime contractor of this project is Philips Research Lab in Eindhoven . Our contribution focuses on the construction of video summaries, and on the automatic structuring of video recordings, to be used in user interface such as PDA. • Semantic Feature Extraction We have participated to the Video Track of the 2003 TREC Conference. The TREC conference (Text Retrieval Conference) contains a series of benchmarks to compare information retrieval systems on selected tasks (called tracks). Since 2001, a Video Track has been organized to establish the state of the art in video searching. We have participated on the task called “Semantic feature extraction” by which video clips have to be classified into one out of several semantic categories such as: indoor, outdoor, cityscape, people, etc… As part of the preparation of those experiments, we have participated to a collaborative effort lead by IBM Research to annotate a set of training video with Mpeg7 semantic attributes.
• Image Graph Matching Investigator: Dr Benoit Huet Based on our previous work on technical image content-based retrieval and indexing, we have initiated a novel approach to indexing and retrieval of video objects based on both geometric and structural information. The objective here is to improve the quality of video indexing, retrieval and summarization by looking at objects (or image regions) within the video instead of the entire frame. Our effort to achieve this follows two separate tracks. The first concentrates on the issue of adapting our previous graph based methods to this new domain. In order to efficiently perform the matching of the complex data structures representing video objects on the very large data volumes required for video analysis, we explore the construction of efficient index structures. The other approach aims at extending our recent work on video object classification using on Latent Semantic Analysis of image regions. The basic technique offers promising results in spite of the fact that the automatically segmented region are characterized some attributes but does not make use of the relationship (connectivity and relative position) between regions (object sub-parts). Our current study aims at incorporating relational information within the object representation and classification (LSA) as well as evaluating the improvement of performance when compared to our basic implementation. • Information Retrieval and Filtering • Personalization for Interactive TV In June 2002, we have started a European project called GMF4iTV (Generic Media Framework for Interactive Television). The prime contractor is Portugal Telecom Inovacao, and the main technical partner is IRT (Institut fur Rundfunktechnik , Germany ). This project aims at implementing a complete chain for TV Programs with interactive objects, from the production side up to the set-top box on the user side. On the production side, an editor workstation will allow to select objects in the video and association to those some annotations (text, graphics, Mpeg4 sequences, web refrerences, etc..). On the set-top box side (based on the MHP platform), the user will be able to select some of the interactive objects and process the associated annotations. In 2003, we have defined the personalization mechanisms to be included in the prototype. Those mechanisms will be based on the comparison of semantic categories for the objects and scenes with the characteristics of the user profile. Based on a set of rules, decisions will be taken on which objects should be activated, and which type of annotation should be displayed to the user. Some simulatied videos have been constructed to illustrate those mechanisms. • Research Team Bernard Mérialdo (Professor) Benoît Huet (Assistant Professor) Itheri Yahiaoui (PhD student) Fabrice Souvannavong (PhD student) Joakim Jiten (PhD Student) Pierre Teller (Master student) Lukas Kohl (Master student) • Contracts and Grants SPATION is a European project in the IST Program. GMF4iTV is a European project in the IST Program. • Selected publications Ithery Yahiaoui, Bernard Merialdo et Benoit Huet, " Comparison of multi-episode video summarisation algorithms ", EURASIP Journal on Applied Signal Processing, Special issue on Multimedia Signal Processing, Vol. 2003, No. 1, page 48-55, January 2003. F. Souvannavong, B. Merialdo and B. Huet, " Video Content Structuration With Latent Semantic Analysis ", Third International Workshop on Content-Based Multimedia Indexing, CBMI 2003, 22-24 September 2003, Rennes , France . F. Souvannavong, B. Merialdo and B. Huet, " Semantic Feature Extraction using Mpeg Macro-block Classification ”, NIST Special Publication: SP 500-251, The Eleventh Text Retrieval Conference (TREC 2002 Video Track), 2003. F. Souvannavong, B. Merialdo and B. Huet, " Classification Sémantique des Macro-Blocs Mpeg dans le Domaine Compressé .", CORESA 2003,16 - 17 Janvier 2003, Lyon France. • Biography Bernard Mérialdo graduated from the Ecole Normale Supérieure (Mathématiques) in 1975. He received a Ph.D. in Computer Science from Paris 6 University in 1979 and an "Habilitation à Diriger des Recherches" from Paris 7 University in 1992. He first taught at the Faculty of Sciences in Rabat ( Morocco ). In 1981, he joined the IBM France Scientific Center in Paris , where he led several research projects on natural language processing and speech recognition using probabilistic models. From 1988 to 1990, he was a visiting scientist in the IBM T.J Watson Research Center in Yorktown Heights , N.Y. ( USA ). In 1992, he joined the Multimedia Communications Department of the Institut Eurécom. His current research topics are multimedia indexing and information filtering applications. He is a member of IEEE, ACM, associate editor for the IEEE Transaction on Multimedia, and general chair for the ACM Multimedia 2002 conference. He is currently Head of the Multimedia Communications Departement at Eurecom, and Director of the CNRS FRE 2660. Speech and Audio processing in Multimedia Application Principal investigator: C.J.Wellekens • Research Themes • Navigation in Speech and Audio Databases (N. de Saint Aubert) In this first year, focus has been put on the search for recognition of large vocabulary. Indeed, one way for indexing is to use standard tools for natural language indexing after having converted speech to text. A report is under preparation. • Speaker adaptation ( N. Annabi , trainee) Nefissa Annabi implemented the MLLR adaptation technique in our recognition Software and conducted severeal tests • Automatic Modeling of Pronunciation variations (K.T.Lee) K.T.Lee wrote his thesis report and presented his defense in September 2003. He is no longer with our laboratory. • Speaker Identification Using Knowledge Inheritance and Eigenvoices (Yassine Mami) Yassine Mami completed his thesis in October and presented his results on speaker identification using their localization in a trained speaker space. • Bayesian Networks and Variational Training ( Fabio Valente) The thesis topic has progressively evolved towards the variational Bayesian training. Several applications have been reported in conference papers and more explicit internal reports. Also experiences were pursued on Valente's master thesis topic devoted to Minimim classification error criteria and eigenvoices applied to speaker identification. A very interesting result due to Variational Bayesian training is the automatic determination of the optimal number of clusters in a speaker space and of the GMM complexity (number of Gaussians). • Improving speech recognition with microphone arrays ( Luca Brayda ) Microphone arrays have been used for several applications like speaker localization, Dereverbaration of speech, speech enhacement… On the other hand, several techniques have been suggested in the last years for multistream (or multiband recognition). The objective of the thesis is to try to use the multichannel information generated by the array in a multistream recognizer. The thesis started mid October 2003. 6th FP Strep DIVINES (Christian Wellekens) A consortium with Multitel (B) (project leader), France Telecom Lannion (F), Eurecom (F), University of Oldenburg (D), Babeltechnologies (B), Loquendo (I), Laboratoire d'Informatique d'Avignon (F), Politecnico di Torino (I) submitted the project DIVINES which was accepted for 3 years. The purpose is to systematically analyze how intrinsic variations in speech impair the recognition. Detailed tests will conducted at the feature level, acoustic modeling, lexical modeling to track the sources of misrecognition. The experience of the University of Oldenburg will guarantee that the tests will be organized taking account of the strict methology of hearing measurements. The originality of the project is its aim to look in the recognition mechanisms instead of proposing new algorithm and just experimenting with the hope of better scores. The project drew the attention of two US researchers (Steven Greenberg ICSI and Rick Rose, Mc Gill University) who will collaborate with US and Canadian grants . One PhD student will be hired at Eurecom as well as a post-doc for the duration of the project. The project will start in February 2004. • Publications • Conferences K.T. Lee,L. Melnar, J. Talley, C.J.Wellekens, " Symbolic Speaker Adaptation with Expanded Phone Inventory ", ICASSP 2003, Hong Kong . F.Valente, C.J.Wellekens, " Minimum classification error /eigenvoices training for speaker identification", ICASSP 2003, 28th International Conference on Acoustics, Speech and Signal Processing - 06-10 April 2003, Hong Kong F.Valente,C.J. Wellekens, " Variational bayesian GMM for speech recognition",
Eurospeech 2003, 8th european conference on speech communication and technology - September 1-4, 2003, Geneva, Switzerland F.Valente, C.J. Wellekens, " Maximum entropy discrimination (MED) feature subset selection for speech recognition" , ASRU 2003, IEEE Automatic Speech Recognition and Understanding Workshop, November 30 -- December 4, 2003, St. Thomas, Virgin Islands, USA • Research Report F.Valente, C.J. Wellekens, " Variational bayesian learning for gaussian mixture models and hidden markov models", Rapport de recherche N° 79 - RR-03-079 F. Valente, C.J. Wellekens, " Variational Bayesian features selection", Rapport de recherche RR-03-087 F.Valente, C.J. Wellekens, " Maximum entropy discrimination (MED) feature subset selection for speech recognition", Rapport de recherche RR-03-086 L. Brayda, " Signal and Model-based Noise Compensation Methods for Automatic Speech Recognition " Master Thesis, Institut Eurecom, sept 2003. K.T Lee, " Dynamic Pronunciation Modeling Using Phonetic Features and Symbolic Speaker Adaptation for Automatic Speech Recognition" , Thèse de doctorat EPFL (Eurecom), September 2003. Y.Mami, " Reconnaissance de locuteurs par localisation dans un espace de locuteurs de référence" , Thèse de doctorat, ENST (Eurecom), October 2003
• Accepted paper F.Valente, C.J.Wellekens, " Variational Bayesian Feature Selection for Gaussian Mixture Models", ICASSP 2004 Montreal (accepted) L. Brayda, L. Rigazio, R. Boman and J-C. Junqua, " Sensitivity Analysis of Noise Robustness Methods" , Panasonic Speech Technology Laboratory and
Institut Eurecom, ICASSP 2004. (accepted) • Research team Nefissa Annabi, Trainee 1/4/03 -1/10/03 Luca Brayda , Ph.D student since 1/10/03 Kyung Tak Lee, Ph.D student , Defense on 17/9/03 Yassine Mami (France Telecom R&D), Ph.D student, Defense on 21/10/03 Nicolas de Saint Aubert, Ph.D student Fabio Valente, Ph.D student Christian Wellekens , Professor • Funding and contract Fabio Valente and Luca Brayda are supported by grants from the MRST (Ministère de la recherche, de la science et de la technologie). Nicolas de Saint Aubert is partly supported by a grant of the PACA ( Provence , Alpes, Cote d'Azur Region) and by Eurecom. Yassine Mami is supported by France Telecon R/D. Nefissa Annabi was supported by Eurecom. • Scientific activities in 2003 Christian Wellekens attended several meetings of ISCA board in Martigny (May) and in Geneva (September). has attended two COST278 meetings in Barcelona (February) and Stockholm (May) has been reviewer for 3 IST programs for HLT European Union in Luxemburg and Trento (Reports). was an invited speaker for the conference Transits in Budapest in February: “Pertinent information: any time, anywhere?” was an invited speaker at the first RIAM conference in Marseille (April 2003): “Naviguer ais\'ement dans les fichiers audio?” is a guest editor for a Special Issue of Speech Communication on Adaptive Methods for Automatic Speech Recognition. was a member of the program committee of Eurospeech 2003 ( Geneva ) is a member of the program committee of ICASSP 2004 participated to the accepted 6 th FP Strep DIVINES. is responsible of the Commission Recherche/Development of Telecom Valley Sophia Antipolis and organized the Prize of Innovation and for the first time, the Prize of the Best Doctoral Thesis (November).
• Biography Christian Wellekens graduated as an E.Mech Engineer in 1965 at the Catholic University of Louvain (UCL). He worked as a development engineer in the development laboratory of High Frequency Systems of MBLE-Philips ( Brussels ) and then joined the Philips Research Lab Brussels where he worked successively in filter design, piezoelectrical device modeling, digital signal processing, speech recognition and neural networks. In 1974 he got his Ph.D. at EPFL under the direction of Prof.Jacques Neyrinck. In 1989-90, he spent a sabbatical year at Bellcore in the department of Prof. Tom Landauer. In 1991, he was appointed Scientific Advisor at Lernout and Hauspie Speechproducts. He left in 1992 to head the Multimedia Communications Dpt of Eurecom till 1998. He organized a workhop in Portugal on Neural Networks in 1990 and an ISCA Technical and Research Workshop on Adaptation Methods in Speech Recognition at Sophia Antipolis in 2001. He is a senior member of IEEE, member of the board of the International Speech Communication Association (ISCA) and member of the Speech Technical Committee of IEEE. Principal Investigator: Jean-Luc D ugelay Our current research is focused on security (watermarking and biometrics) and facial image analysis for person authentication, talking heads and human machine interface. Much of our work can be found at http://www.eurecom.fr/~image. • Highlights- Co-organization with Panasonic Research and UC, Santa Barbara of a workshop on Biometrics; - Best student paper at IEEE ICIP 2003 (F. Perronnin); - 2 Ph.D. defenses obtained with highest honors: - C. Rey on digital watermarking; - A. Andrés on facial expressions. - Core partner of a new major European Network of Excellence on Multimodal Interfaces (SIMILAR). • Research Themes • Biometrics The security field uses three different types of authentication: something you know, something you have, or something you are - a biometric. Common physical biometrics include fingerprints, hand geometry; and retina, iris, or facial characteristics. Behavioral characters include signature, voice. Ultimately, the technologies could find their strongest role as interwined and complementary pieces of a multifactor authentication system. In the future biometrics is seen playing a key role in enhancing security, residing in smart cards and supporting personalized Web e-commerce services. Personalization through person authentication is also very appealing in the consumer product area. Florent Perronnin (previously research eng. at PANASONIC Research Santa Barbara) joined our team in November 2001. His Ph.D. dissertation is dedicated to image processing for biometrics. In collaboration with UC, Santa Barbara (Prof. Ken Rose), a new efficient probabilistic framework to model relations between different images and to provide a matching score has been designed. The proposed approach is based on local flexible mappings coupling with a global supervision. Competitive results have been obtained for face recognition (robust to variations due to facial expressions). F. Perronnin received the IEEE ICIP 03 Best Student Paper Award. This work is partially done in collaboration with France Télécom Research. • Watermarking Image Watermarking is now a major domain. Basically, watermarking allows owners or providers to hide an invisible and robust message inside a digital Multimedia document, mainly for security purposes, in particular, owner or content authentication. There is a complex trade-off between the different parameters: capacity, visibility and robustness. Christian Rey defended his thesis on February 2003 on robust image watermarking. His main contribution was on the pre-formatting of the watermark using turbo codes (in collaboration with ENST Bretagne) and reference bits for self-synchronisation. The proposed technique includes a high level of robustness, in particular against local geometric attacks (i.e. cracker Stirmark). He obtained his Ph.D. with highest honors. His thesis has been labeled by Telecom Valley . Most existing methods for video watermarking are based on a frame-by-frame approach; more sophisticate and powerful frameworks have to be investigated. Gwenaël Doërr, previously trainee at NECI Princeton, joined our team in November 2001. He is working on video watermarking, with a special emphasis on the problem of temporal collusion (attacks and counter-measures) Yann Bodo, Ph. D. student at France Télécom Research Rennes, is mainly working on waterscrambling and on watermarking of MPEG steams via the motion vectors. Emmanuel Garcia joined our team for an internship in January 2000. He is currently working on 3-D video object watermarking and data hiding, in particular for 3-D faces. An original approach based on texture has been proposed and developed. He is contributing to the project Semantic-3D. • Facial Image Analysis (pose & expressions) Ana Cristina Andrés Del Valle previously trainee at AT&T Research, and originally student from UPC, joined our group in January 2000 and defended her thesis on facial expressions on September 2003. Her main contribution was to propose a complete framework to analyse ocular expressions (eyes and eyebrows) robust to head pose variations. In addition to face cloning, new applications are expected in the domains related to human machine interface. • Wireless Multimedia Imaging Caroline Mallauran joined our team in October 2001, in collaboration with the dept. of Mobile Communications. She was mainly working on a new application of video spatialization (i.e. visual navigation) via mobile communications channels, i.e.: Wireless Multimedia Imaging. This work was part of a French project sponsored by the RNRT (National Network for Research in Telecommunications). She left Eurécom in April 2003. • Research Team Jean-Luc DUGELAY (Professor) Emmanuel Garcia (Ph.D. student, Univ. of Nice ) Florent Perronnin (Ph.D. student, EPFL) Gwenaël Doërr (Ph. D student, Univ. of Nice ) Ana Cristina Andrés Del Valle (Ph.D. student, ENST) Christian Rey (Ph.D. Student, Univ. of Avignon ) Caroline Mallauran (Reasearch Eng. ) Lucas Grange (trainee, University of Nice ) Vanessa Lopez (trainee, UPC) Yann Bodo (external Ph.D Student) • Contracts and Grants PLATON is a National project SEMANTIC-3D is a National project SIMILAR is a European project The Ph.D. of E. Garcia is partially supported by a DEA thesis grant. The Ph.D. of G. Doerr is partially supported by a DEA thesis grant. The Ph.D. of F. Perronnin on Biometrics is partially supported by France Télécom Research. The Ph.D. of A. Andrés on Face cloning was partially supported by France Télécom Research. • Selected publication F. Perronnin, J.-L. Dugelay and K. Rose, " Deformable face mapping for person identification ", ICIP, 2003, vol. 1, pp. 661-664. E. Garcia, J.-L. Dugelay, " Texture-Based Watermarking of 3D Video Objects ",
IEEE Transactions on Circuits and Systems for Video Technology, Vol. 13, No. 8, August 2003 G. Do ë rr and J.-L. Dugelay , A Guide Tour of Video Watermarking .
In " Signal Processing: Image Communication " , 18(4):263-282, 2003. G. Doërr and J.-L. Dugelay , "New Intra-Video Collusion Attack Using Mosaicing ."
In Proceedings of the IEEE International Conference on Multimedia & Expo , vol. II, pp. 505-508, 2003. C. Rey, K. Amis, J.-L. Dugelay, R. Pyndiah & A. Picart, " Enhanced robustness in image watermarking using block turbo codes", Proceedings of SPIE, Electronic Imaging, Security and Watermarking of Multimedia Contents , Janvier 2003, San-Jose, CA A. C. Andrés del Valle & J.-L. Dugelay, " Making Machines Understand Facial Motion and Expressions like Humans Do "
June 21st-27th, 2003 Crete - Grece , HCI International 2002 Bodo, Yann; Laurent, Nathalie;Dugelay, Jean-Luc, " Watermarking video, hierarchical embedding in motion vectors", ICIP'2003, IEEE International conference on image processing, September 14-17, 2003 , Barcelona , Spain Ana Andres, " Facial motion analysis on monocular images for telecom applications:coupling expression and pose understanding", PhD student's thesis September 2003, Telecom Paris Christian Rey, " Watermarking – Increased robustness and image integrity", PhD student's thesis, February 2003, University of Avignon , France . • Biography Jean-Luc Dugelay received the Ph.D. degree in Computer Science in 1992 from the University of Rennes . Doctoral research was carried out, from 1989 to 1992, at the France Telecom Research Laboratory in Rennes (formerly CNET - CCETT). He then joined the Institut Eurécom (Sophia Antipolis), where he is currently a Professor in the Department of Multimedia Communications. His research interests are in the area of multimedia signal processing and communications; including security imaging (i.e., watermarking and biometrics), image/video coding, facial image analysis, virtual imaging, face cloning and talking heads. He is an author or coauthor of more then 65 publications that have appeared as journal papers or proceeding articles, 3 book chapters, and 3 international patents. He gave several tutorials on digital watermarking and image compression at major conferences. He has been an invited speaker and/or member of the program committee of several scientific conferences and workshops. Jean-Luc Dugelay is an Associate Editor for several major international journals, and an active member of the IEEE Signal Processing Society. Affective Social Computing Principal Investigator: Prof. Christine LISETTI Prof Christine Lisetti joined Eurecom in September 2003. • Research Themes The overall goals of the Affective Social Computing Group are to enhance human-computer interaction (HCI) by enriching it with affective knowledge. Building upon the PI's previous research, the work on affective social computing will expand different research themes, focusing on computational models of affective social cognitive intelligence by considering the notion of social expertise in terms of: (1) external communicative behavior and (2) internal motivational goal-based abilities. • Multimodal Affective User Interfaces We are studying the role of emotions in communication, designing and developing computer systems and interfaces with increased "awareness" of their user's affective states and communication patterns in order to adapt appropriately to their users depending upon the application. Building upon the PI's previous results, we will investigate different ways that the interpretation of multimodal physiological signals associated with emotional states can be visualized and responded to with the use of anthropomorphic agents and expressive graphical avatars having associated internal affective-cognitive states. • Emotion-Based Architectures for Socially Intelligent Agents We are studying mechanisms to model computationally affective social cognitive intelligence to enhance both agents' autonomy, and human-computer or human-robot interaction. Our approach will involve the design of an affective-cognitive hybrid architecture which, building upon the PI's previous research, will synthesize affective-cognitive states for artificial agents. We will study ways to combine the reactive behavioural approach with the deliberative planning BDI approach (of modelling cognition in terms of Beliefs, Desires, and Intentions). • Research Team Professor Lisetti is actively searching qualified students to start her new research group. • Research Funding Actions Taken Professor Lisetti submitted various research funding grant applications. • Selected publications C. Lisetti. *(Invited Key Note). What Do Socially Intelligent Agents Need? To Appear In Proceedings of the Cognitive Informatics National Conference, ( Montreal , CANADA , May 2004) . C. Lisetti (accepted with revisions). Building a Model of User's Emotions. AI Communications The European Journal on Artificial Intelligence . A. Cami, C. Lisetti and M. Sierhuis (accepted). Simulating the Multi-Level Process of Human Emotions in Multi-Agent Systems. To Appear In Proceedings of the 17 th European Meeting on Cybernetics and Systems Research EMCSR 2004 Symposium on Affective Computational Entities , ( Vienna , Austria , April 2004). C. Lisetti, and A. Marpaung (accepted). A 3-Layered Emotion-Based Architecture for a Social Robot. To Appear In Proceedings of the 17 th European Meeting on Cybernetics and Systems Research EMCSR 2004 Symposium on Affective Computational Entities , ( Vienna , Austria , April 2004). A. Cami, C. Lisetti and M. Sierhuis (accepted). Toward the Simulation of a Multi-Level Model of Human Emotions. To Appear In Proceedings of the AAAI Spring Symposium Series on Architectures for Simulating Emotions , ( Stanford , CA , March 2004). • Conference Committees • Biography Christine Lætitia Lisetti is a Professor at the Institut EURECOM in the Multimedia Communications Department. Previously, she was an Assistant Professor in Computer Science at the University of Central Florida , USA . From 1996 to 1998, she was a Post-Doctoral Fellow at Stanford University in the Department of Computer Science and the Department of Psychology jointly. She received a B.S. in computer science cum laude , an M.S. and a Ph.D. in computer science in 1988, 1992, and 1995 respectively, from Florida International University . She has won multiple awards including the National Institute of Health Individual Research Service Award (USA), the AAAI Nils Nilsson Award for Integrating AI Technologies , and the University of Central Florida COECS Distinguished Research Lecturer Award . Dr. Lisetti has focused her research interests on the study of emotional intelligence and computational models of affective processes. Her research involves the use of HCI and artificial intelligence (AI) techniques to model affective knowledge computationally. Her research has been granted support from federally funded agencies such as the National Institute of Health , the Office of Naval Research , and US Army STRICOM and also from industries such as Interval Research Corporation and Intel Corporation . She is regularly invited to serve on program committees of international conferences, and has co-chaired several international events on Affective Computing. She is a member of the IEEE , ACM , and AAAI . | 
