Model Fitting in Computer VisionAbstract:
Many computer vision problems can be formulated as data fitting. In multi-class multi-instance fitting, the input data is interpreted as a mixture of noisy observations originating from multiple instances of multiple model types, e.g. as k lines and l circles in 2D edge maps, as k planes, l cylinders and m point clusters in 3D laser scans, as multiple homographies or fundamental matrices consistent with point correspondences in multiple views of a non-rigid scene.
I will review the basic assumptions, strenghts and weakness of two historially popular data fitting methods - RANSAC and the Hough transform. I will then present a novel method, called Multi-X, for general multi-class multi-instance model fitting. The proposed approach combines a random sampling strategy like RANSAC, lobal energy minimization using alpha-expansion, and mode-seeking in the parameter domain like the Hough Transform. Multi-X outperforms significantly the state-of-the-art on standard datasets, runs in time approximately linear in the number of data points, an order of magnitude faster than available implementations of commonly used methods.
Jiri Matas is a full professor at the Center for Machine Perception, Czech Technical University in Prague. He holds a PhD degree from the University of Surrey, UK (1995). He has published more than 200 papers in refereed journals and conferences. His publications have approximately 34000 citations registered in Google Scholar and 13000 in the Web of Science. His h- index is 65 (Google scholar) and 43 (Clarivate Analytics Web of Science) respectively.
He received the best paper prize e.g. at the British Machine Vision Conferences in 2002 and 2005, at the Asian Conference on Computer Vision in 2007 and at Int. Conf. on Document analysis and Recognition in 2015. J. Matas has served in various roles at major international computer vision conferences (e.g. ICCV, CVPR, ICPR, NIPS, ECCV), co-chairing ECCV 2004, 2016 and CVPR 2007. He is on the editorial board of IJCV and was the Associate Editor-in-Chief of IEEE T. PAMI. He served on the computer science panel of ERC.
His research interests include visual tracking, object recognition, image matching and retrieval, sequential pattern recognition, and RANSAC- type optimization metods.
Deep Learning and Kernel Machines: towards a Unifying Framework.Abstract:
Among powerful data-driven methods in machine learning are deep learning & neural networks and support vector machines & kernel methods. Deep architectures have catched major attention with convolutional neural networks, stacked autoencoders, deep Boltzmann machines and generative adversarial networks. Support vector machines on the other hand were overcoming the issues of non-convexity occuring in the training of neural networks. Moreover it has stimulated the wide use of kernel-based approaches for different tasks in supervised, unsupervised and semi-supervised learning.
In this talk we show new synergies between deep learning, neural networks, least squares support vector machines and kernel methods. An important role at this point is played by different duality principles, paving the way towards a unifying framework.
Johan A.K. Suykens received the master degree in Electro-Mechanical Engineering and the PhD degree in Applied Sciences from the Katholieke Universiteit Leuven, in 1989 and 1995, respectively.
He has been a Postdoctoral Researcher with the Fund for Scientific Research FWO Flanders and is currently a full Professor with KU Leuven. He is author of the books "Artificial Neural Networks for Modelling and Control of Non-linear Systems" (Kluwer Academic Publishers) and "Least Squares Support Vector Machines" (World Scientific), co-author of the book "Cellular Neural Networks, Multi-Scroll Chaos and Synchronization" (World Scientific) and editor of the books "Nonlinear Modeling: Advanced Black-Box Techniques" (Kluwer Academic Publishers), "Advances in Learning Theory: Methods, Models and Applications" (IOS Press) and "Regularization, Optimization, Kernels, and Support Vector Machines" (Chapman & Hall/CRC).
In 1998 he organized an International Workshop on Nonlinear Modelling with Time-series Prediction Competition. He has served as associate editor for the IEEE Transactions on Circuits and Systems (1997-1999 and 2004-2007), the IEEE Transactions on Neural Networks (1998-2009) and the IEEE Transactions on Neural Networks and Learning Systems (from 2017). He received an IEEE Signal Processing Society 1999 Best Paper Award and several Best Paper Awards at International Conferences. He is a recipient of the International Neural Networks Society INNS 2000 Young Investigator Award for significant contributions in the field of neural networks. He has served as a Director and Organizer of the NATO Advanced Study Institute on Learning Theory and Practice (Leuven 2002), as a program co-chair for the International Joint Conference on Neural Networks 2004 and the International Symposium on Nonlinear Theory and its Applications 2005, as an organizer of the International Symposium on Synchronization in Complex Networks 2007, a co-organizer of the NIPS 2010 workshop on Tensors, Kernels and Machine Learning, and chair of ROKS 2013. He has been awarded an ERC Advanced Grant 2011 and 2017, and has been elevated IEEE Fellow 2015 for developing least squares support vector machines.
Casting Data for Deep Speech and Audio AnalysisAbstract:
Deep Learning and many Machine Learning problems often suffer from sheer endless cravings for training data. This holds in particular for Speech and general Audio Analysis beyond Speech Recognition where data is usually sparse. Furthermore, it is often difficult to label such data as in subjective annotation tasks like speaker emotion or likability, where potentially several annotations are needed per audio segment resulting in high labelling costs and efforts. To ease this omnipresent and dominating bottleneck, this talk presents avenues towards highly efficient “casting” of training data. This includes synthesis of data such as by Generative Adversarial Networks on feature, but also on audio level aiming at overcoming model collapse to reach high diversity of generated data. Furthermore, rapid and targeted exploitation of social media such as YouTube by small-world modeling of recommendations combined with unsupervised content verification is introduced. Other avenues shown include transfer learning – even from other modalities such as from the computer vision domain – and advanced forms of collaborative learning based on dynamic active learning where also annotators are modelled for increased efficiency. Results are presented across a variety of Computer Audition tasks for speech and sound analysis.
Björn W. Schuller is Full Professor & Head of the Chair of Embedded Intelligence for Health Care and Wellbeing, University of Augsburg, Germany, Professor of Artificial Intelligence & Head of GLAM - Group on Language, Audio & Music, Imperial College London/UK, Chief Scientific Officer (CSO) and Co-Founding CEO, audEERING/Germany, and Visiting Professor, Harbin Institute of Technology/P.R. China. He received his diploma, doctor in engineering, and habilitation from TUM/Germany.
Björn co-authored more than 750 peer-reviewed technical contributions (20,000+ citations, h-Index 68). He is a Fellow of the IEEE, President-Emeritus of the AAAC, Editor-in-Chief of the IEEE Transactions on Affective Computing, and repeated General Chair (ACII 2019, ICMI 2014) and Program Chair (INTERSPEECH 2019, ICMI 2019, ACII 2015, ICMI 2013, SocialCom 2012, and ACII 2011) amongst manifold further service to the community.
Big Data: The necessary triplet of technology, quality data and scalable algorithmsAbstract:
Big data applications are emerging during the last years, and researchers from many disciplines are aware of the high advantages related to the knowledge extraction from this type of problem. The term big data is classically associated with the technologies (Hadoop, Spark, Flink,.), together with the well known characteristics of volume, variety and velocity of the data. Beyond these classic concepts, it is necessary to discuss about other essential elements to tackle a big data project, to have quality data and design scalable algorithms for analysis, efficient and effective algorithms. In this talk I will present big data from this triplet, that must be complemented, technology, data and algorithms. We will pay special attention to the last two elements. The first one, the transformation of the original data into quality data (smart data) providing value and knowledge. Second, the design of the algorithms, new scalable distributed machine learning algorithms under the MapReduce framework. We will include some cases of study and research challenges.
Francisco Herrera is currently a Full Professor in the Department of Computer Science and Artificial Intelligence at the University of Granada.
Francisco Herrera (SM'15) received his M.Sc. in Mathematics in 1988 and Ph.D. in Mathematics in 1991, both from the University of Granada, Spain. He is currently a Professor in the Department of Computer Science and Artificial Intelligence at the University of Granada. He has been the supervisor of 43 Ph.D. students. He has published more than 400 journal papers that have received more than 65000 citations (Scholar Google, H-index 128). He is coauthor of the books "Genetic Fuzzy Systems" (World Scientific, 2001) and "Data Preprocessing in Data Mining" (Springer, 2015), "The 2- tuple Linguistic Model. Computing with Words in Decision Making" (Springer, 2015), "Multilabel Classification. Problem analysis, metrics and techniques" (Springer, 2016), "Multiple Instance Learning. Foundations and Algorithms" (Springer, 2016), "Learning from Imbalanced Data Sets" (Springr 2018).
He currently acts as Editor in Chief of the international journals "Information Fusion" (Elsevier) and "Progress in Artificial Intelligence (Springer). He acts as editorial member of a dozen of journals.
He received the following honors and awards: ECCAI Fellow 2009, IFSA Fellow 2013, 2010 Spanish National Award on Computer Science ARITMEL to the "Spanish Engineer on Computer Science", International Cajastur "Mamdani" Prize for Soft Computing (Fourth Edition, 2010), IEEE Transactions on Fuzzy System Outstanding 2008 and 2012 Paper Award (bestowed in 2011 and 2015 respectively), 2011 Lotfi A. Zadeh Prize Best paper Award of the International Fuzzy Systems Association, 2013 AEPIA Award to a scientific career in Artificial Intelligence, and 2014 XV Andalucía Research Prize Maimónides (by the regional government of Andalucía), 2017 Security Forum I+D+I Prize, 2017 Andalucía Medal (by the regional government of Andalucía) and 2018 "Granada. City of Science and Innovation". He has been selected as a Highly Cited Researcher http://highlycited.com/ (in the fields of Computer Science and Engineering, respectively, 2014 to present, Clarivate Analytics). His current research interests include among others, soft computing (including fuzzy modeling, evolutionary algorithms and deep learning), computing with words, information fusion and decision making, and data science (including data preprocessing, prediction and big data).