Title: "Anomaly Detection with Sparse Representations" Abstract: Sparse representations have shown to be a key ingredient for achieving state-of-the-art performance in several imaging and signal processing applications. Models yielding sparse representations were originally introduced for descriptive tasks such as denoising/regression, but more recently discriminative tasks, classification in particular, have also been addressed in the sparse representation literature. During this talk, I will present an anomaly detection framework based on sparse representation. The underlying assumption is that data generated in nominal (stationary) conditions admit a sparse representation w.r.t. a suitable dictionary, thus that they are well approximated in a union of low-dimensional subspaces. Anomalies, i.e. data that do not conform with nominal conditions, are detected when they depart from these subspaces. I will illustrate the effectiveness of this approach in two application scenarios. The first one consists of an industrial monitoring application, where a scanning electron microscope (SEM) is used to supervise the production of nanofibres: anomalous patterns indicate a quality degradation in the produced materials, and as such have to be automatically detected. The second one consists of an environmental monitoring application, where nodes of a sensor network acquire acoustic emissions (bursts) to monitor a rock face. The stream of bursts is then analyzed to detect structural changes which might be associated with macroscopic evolutions of the monitored phenomenon, eventually leading to a rock collapse. BioSketch Giacomo Boracchi received the M.S. degree in Mathematics from the Universita` Statale degli Studi di Milano, Italy, and the Ph.D. degree in Information Technology at Politecnico di Milano, Italy, in 2004 and 2008, respectively. He was researcher at Tampere International Center for Signal Processing, Finland, during 2004-2005. Currently, he is an assistant professor at the Dipartimento di Elettronica, Informazione e Bioingegneria of the Politecnico di Milano. His main research interests are mathematical and statistical methods for image analysis/processing as well as learning techniques for dynamic and nonstationary environments.