February 26th, 2020

Dear All,

IEEE Signal Processing Society, Bangalore Chapter and Department of Computational and Data Sciences,Indian Institute of Science Invite you to the following talk:

SPEAKER : Badri Narayana Patro, Postdoctoral Fellow, Google Research India
TITLE : “Towards Understanding Vision and Language Systems: Controllability, Uncertainty and
Interpretability for VQA and VQG”
Venue : #102, CDS Seminar Hall
Date & Time : Feb 26, 2020, 04:00 PM

Intelligent interaction between humans and automated systems is an important area of research in computer vision, natural language processing, and machine learning. One such interaction involves visual question answering (VQA), where questions asked by humans are answered by a machine. This involves using information from different modalities, such as image and language. Such automated systems have many application possibilities in building assistive technology for visually impaired people and in efficient surveillance and robotics systems. Now to have a healthy interaction, it is important also to have engaging questions being generated by machines. Generating natural questions based on an image is a challenging semantic task termed visual question generation (VQG) that requires multimodal representations. Images can have multiple visual and language contexts that are relevant for generating questions, namely places, captions, and tags. In the literature, there are several ways based on deep learning to solve these problems. In this talk, we aim to understand these techniques better to incorporate controllability, obtain an estimate of the uncertainty of solving the problem, and also be able to explain these techniques using visual and textual explanations in a robust manner.

Badri Narayana Patro is currently a Postdoctoral Fellow at Google Research Lab for AI, India. He has submitted his PhD thesis on the title “ Towards Understanding Vision and Language Systems” in the Department of Electrical Engineering from Indian Institute of Technology, Kanpur. He received his M.Tech degree in Department of Electrical Engineering from Indian Institute of Technology, Bombay and received his B.Tech degree in Electronic and telecommunication Engineering from National Institute of Science and Technology, Odisha. He was working as a Lead engineering in Samsung R&D institute Delhi, India. He has also worked for Harman International limited, Pune, India as an associate software engineer and assistant software engineer at Larsen and Toubro at Mysore. His expertise is in the fields of Computer Vision, Natural Language Processing, Pattern Recognition and applied Machine Learning. He has authored papers in different venues such as CVPR, ICCV, AAAI, EMNLP, COLING, WACV. He works the Vision and Language team at Google Research Lab and works closely with allied teams such as AI for Social Goods and Google Image teams within Google. He has served as a reviewer for CVPR, ICCV, AAAI, ACL, TIP, TMM, WACV, ICVGIP, NCVPRIPG conferences and Journals. He is a member of the ACL, AAAI, and CVF. He actively collaborates with institutes to further his research interests.

Host Faculty: Prof. Venkatesh Babu

February 4th, 2020
Title: Decomposition of dynamical networks with tensor decomposition
Speaker: Dr. Pierre Borgnat (Director of research, LPENSL, CNRS, ENS Lyon, France; Directeur de l’IXXI (Complex Systems Institute Rhône-Alpes))
Venue: Golden Jubilee Seminar Hall, ECE Dept., IISc
Schedule: 04/02/2020, 2.30-3.30 pm
Coffee/tea: 3.30pm
Organized by: Dept. of ECE, IEEE-IISc Student Branch, Signal Processing Society Chapter
Abstract: Dynamical networks are an instance of data that ate studied in many domains, e.g. transportation, social and economical studies, communication networks or biological networks such as brain activity. The range of dynamical features that exist is really wild and many methods have been proposed to extract a reduced number of components, jointly in time and across the (evolving) graph topology. The purpose of the talk is to present some of the works we did in this direction in the context of neuroscience studies, using a novel tensor decomposition approach followed by clustering to extract components representative of various activities in time. The
application context is the study of Functional connectivity (FC) of EEG, which is a graph-like data structure commonly used by neuroscientists to study the dynamic behavior of brain activity. We will show in examples how the proposed approach allows us to decompose data of EEG brain activity of patients suffering from epilepsy, allowing us to infer network components corresponding to the different stages of an epileptic seizure.
Joint work with G. Frusque, P. Gonçalves, J. Jung, R. Cazabet, R. Hamon
Biography: Pierre Borgnat is a CNRS Senior Scientist, at the Laboratory of Physics, ENS de Lyon. Born in France in 1974, he received the Ph.D. degree in Physics and Signal Processing in 2002, and the HDR in 2014. He was a CNRS Chargé de Recherche since 2004, Directeur de Recherche since 2016. He is director of IXXI (Complex System Institute of Rhône-Alpes) since 2014. His research interests are in statistical signal processing, and especially in graph signal processing. He also has interests in complex networks, nonstationary signals or scaling phenomena and machine learning. He works on methodological developments in these domains and studies also several applications of these signal processing methods: Internet traffic modeling and measurements, data processing for physics, analysis of social data, transportation studies and in Neuroscience. He was Associate Editors of IEEE Trans. on Signal Processing (2015-2019) and is currently Area Editor of the same Transactions.
All are Welcome!
February 4th, 2020
IEEE-IISc Student Branch
Signal Processing Society Chapter
Technical talk
Deep Learning for multivariate fractal texture synthesis:
Does it work? How do we know that it works?
Dr. Patrice Abry (IEEE Fellow)
Director of research, CNRS, ENS Lyon, France
Golden Jubilee Seminar Hall
ECE Dept., IISc
06/02/2020, 2.30-3.30 pm
All are Welcome
Coffee/tea 3.30pm
Abstract: Deep Convolutional Generative Adversarial Networks (DCGAN) have been widely used to
synthesize images. Their use remains however concentrated on geometrical images (such as
faces) and they have been much less used for the synthesis of textured images. Our aim is to
investigate the potential of DCGAN to generate multivariate textures. To that end, we make use of a
large set of synthetic multivariate multifractal textures, which consists of a collection of scale-free
(or fractal) textures with non-trivial cross-dependencies (cross-selfsimilarity and cross-
multifractality) to train a DCGAN. We make use of wavelet transforms and wavelet-leaders to
compare the quality of the DCGAN synthesized textures against those of the original textures. We
discuss reproductiblity and convergence issues. Joint work with : V. Mauduit, S. Roux, E.
Quemener, ENS Lyon, France
Patrice Abry (IEEE Fellow) was born in Bourg-en-Bresse, France in 1966. He received the degree of Professeur-Agr eg e de Sciences Physiques, in 1989 at Ecole Normale Sup erieure de Cachan and
completed a PhD in Physics and Signal Processing, at Université Claude-Bernard University in Lyon in 1994. He is a CNRS Senior Scientist, at the Physics dept. of Ecole Normale Superieure de Lyon,
where he is in charge of the Signal, systems and Physics research team. Patrice Abry received the AFCET-MESR-CNRS prize for best PhD in Signal Processing for the years 93-94 and has been elected IEEE Fellow in 2011. He is the author of a book in French dedicated to wavelet, scale invariance and hydrodynamic turbulence and is also the coeditor of a book entitled “Scaling, Fractals and Wavelets”. He has been elected IEEE fellow in 2011 and he serves as an elected member of the IEEE SPS Signal Processing Theory and Methods Technical Committee. His current research interests include wavelet-based analysis and modeling of statistical scale-free dynamics (self-similarity, stable processes, multi-fractal, 1/f processes, long-range dependence, local regularity of processes, infinitely divisible cascades, departures from exact scale invariance). Beyond theoretical developments and contributions in multifractal analysis and stochastic process design. Patrice Abry shows a strong interest into real-world applications, such as hydrodynamic turbulence, computer network teletraffic, heart rate variability.
January 19th, 2020


Venue: Golden Jubilee Seminar Hall, ECE Dept, IISc, Bangalore, India

Time: 4.00 pm On 20th January 2020

Organized by

IEEE EMBS Bangalore Chapter, EMB IISc and RIT Student Chapters,

       IEEE SPS Bangalore Chapter, Department of ECE, Indian Institute of Science.


Title of the Talk: Model-based Signal Processing in Neurocritical Care

Abstract: Large volumes of heterogeneous data are now routinely collected and archived from patients in a variety of clinical environments, to support real-time decision-making, monitoring of disease progression, and titration of therapy. This rapid expansion of available physiological data has resulted in a data-rich – but often knowledge-poor – environment. Yet the abundance of clinical data also presents an opportunity to systematically fuse and analyze the available data streams, through appropriately chosen mathematical models, and to provide clinicians with insights that may not be readily extracted from visual review of the available data streams. In this talk, I will highlight our work in model-based signal processing for improved neurocritical care to derive additional and clinically useful information from routinely available data streams. I will present our model-based approach to noninvasive, patient-specific and calibration free estimation of intracranial pressure and will elaborate on the challenges of (and some solutions to) collecting high-quality clinical data for validation.

Speaker: Prof Thomas Heldt
Massachusetts Institute of Technology, United States   

Thomas Heldt studied physics at Johannes Gutenberg University, Germany, at Yale University, and at MIT. He received the PhD degree in Medical Physics from MIT’s Division of Health Sciences and Technology and undertook postdoctoral training at MIT’s Laboratory for Electromagnetic and Electronic Systems. Prior to joining the MIT faculty in 2013, Thomas was a Principal Research Scientist with MIT’s Research Laboratory of Electronics. He currently holds the W.M. Keck Career Development Chair in Biomedical Engineering. He is a member of MIT’s Institute for Medical Engineering and Science and on the faculty of the Department of Electrical Engineering and Computer Science.

Thomas’s research interests focus on signal processing, mathematical modeling and model identification in support of real-time clinical decision making, monitoring of disease progression, and titration of therapy, primarily in neurocritical and neonatal critical care. In particular Thomas is interested in developing a mechanistic understanding of physiologic systems, and in formulating appropriately chosen computational physiologic models for improved patient care. His research is conducted in close collaboration with clinicians from Boston-area hospitals, where he is integrally involved in designing and deploying high-quality data-acquisition systems and collecting clinical data. 

January 8th, 2020








Title: Towards Autonomous Video Surveillance


Speaker: Prof. Janusz Konrad, Boston University


Time: 1630-1730 hrs (coffee/tea at 4.15pm)


Date:  Thursday, 16 Jan 2020


Venue: Golden Jubilee Seminar Hall, Dept. of ECE, Indian Institute of Science Bangalore



It is estimated that in 2014 there were over 100 million surveillance cameras in the world. Fueled by security concerns, this number continues to steadily grow. As monitoring of video feeds by human operators is not scalable, automatic surveillance tools are needed. In this talk, I will cover a complete video surveillance processing chain, developed over years at Boston University, from low-level video analysis to summarization of dynamic events. I will focus on three fundamental questions posed in video surveillance: “How to detect anomalous events in a visual scene? How to classify those events? How to represent them succinctly?’’ First, I will present “behavior subtraction’’, an extension of “background subtraction’’ to scenes with dynamic backgrounds (e.g., water surface that is notoriously difficult to handle), which can detect complex anomalies in surveillance video. Then, in order to classify activities within the detected anomalies, I will discuss activity recognition on covariance manifolds. Finally, I will describe “video condensation’’, a computational method to succinctly summarize activities of interest for efficient evaluation by human operators.



Janusz Konrad received Master’s degree from Technical University of Szczecin, Poland in 1980 and PhD degree from McGill University, Montréal, Canada in 1984. He joined INRS-Télécommunications, Montréal as a post-doctoral fellow and, since 1992, as a faculty member. Since 2000, he has been on faculty at Boston University. He is an IEEE Fellow and a recipient of several IEEE and EURASIP Best Paper awards. He has been actively engaged in the IEEE Signal Processing Society as a member of various boards and technical committees, as well as an organizer of conferences. He has also been on editorial boards of various EURASIP journals. His research interests include video processing and computer vision, stereoscopic and 3-D imaging and displays, visual sensor networks, human-computer interfaces, and cybersecurity.




Lecture Flyer

January 1st, 2020
IEEE Signal Processing Society, Bangalore Chapter and Department of Computational and Data Sciences, 
Indian Institute of Science Invite you to the following talk:
SPEAKER   : Suraj Srinivas, PhD Scholar, Idiap Research Institute and EPFL, Switzerland
TITLE          : “Neural Network Interpretability using Full-Gradient Representation”
Venue            :  #102 CDS Seminar Hall
Date & Time :  Jan  07, 2020, 12:00 Noon


In this talk, I will introduce a new tool for interpreting neural net responses, namely full-gradients, which decomposes the neural net response into input sensitivity and per-neuron sensitivity components. This is the first proposed representation which satisfies two key properties: completeness and weak dependence, which provably cannot be satisfied by any saliency map-based interpretability method. For convolutional nets, we also propose an approximate saliency map representation, called FullGrad, obtained by aggregating the full-gradient components. We experimentally evaluate the usefulness of FullGrad in explaining model behaviour with two quantitative tests: pixel perturbation and remove-and-retrain. Our experiments reveal that our method explains model behaviour correctly, and more comprehensively than other methods in the literature. Visual inspection also reveals that our saliency maps are sharper and more tightly confined to object regions than other methods. This talk is based on our recent NeurIPS 2019 paper titled “Full-Gradient Representation for Neural Network Visualization”.

Suraj Srinivas is a 3rd year PhD research assistant at Idiap Research Institute and EPFL, Switzerland. He works with Prof. Francois Fleuret on analytically understanding deep learning architectures. Before that, he completed his M.Sc.(Engg.) at CDS, IISc, where he worked with Prof. Venkatesh Babu on neural network compression. His research interest are broadly relating to the robustness, adaptability and interpretability of deep neural networks.

Host Faculty: Prof. Venkatesh Babu

January 1st, 2020

IEEE Signal Processing Society, Bangalore Chapter

Department of Electrical Engineering
Indian Institute of Science (IISc), Bangalore
invite you to the following talk
Title: Utilizing Real-time MRI to Investigate Speech Articulation Disorders
Date and time: January 10, 2020; 11:30AM (Coffee will be served at 11:15AM)
Venue: Multimedia Classroom, Department of Electrical Engineering, IISc.
Speaker: Christina Hagedorn, PhD, SLP, CCC-SLP, City University of New York – College of Staten Island
Over the past two decades, real-time Magnetic Resonance Imaging (rtMRI), elaborating traditional medical MRI, has played a critical role in studying a variety of biological movement patterns. Through collaboration between engineers and speech scientists, rtMRI technology has been applied to the study of speech production. Through semi-automatic detection of air-tissue boundaries and estimation of articulatory kinematics using pixel intensity time functions, rtMRI can be used to quantitatively analyze speech production patterns in both typical and disordered populations. In this work, rtMRI is demonstrated to shed light on aspects of speech produced by individuals with tongue cancer and individuals with Apraxia of Speech that would not be possible using tools providing more limited spatiotemporal information about vocal tract shaping.
Biography of the speaker:
Christina Hagedorn is an assistant professor of Linguistics and director of the Motor Speech Laboratory at the City University of New York (CUNY) – College of Staten Island. Her research focuses primarily on disordered speech production. Her work aims to shed light on the precise nature of articulatory breakdowns in disordered speech and how this can inform theories of unimpaired speech production, as well as lead to refinement of the therapeutic techniques used to address these speech deficits.

She received her Ph.D. in Linguistics from the University of Southern California, where she was a member of the Speech Production and Articulation kNowledge (SPAN) Group, the USC Phonetics and Phonology Group, and was a Hearing and Communication Neuroscience pre-doctoral fellow. She received her clinical training in Communicative Sciences and Disorders at New York University, and holds a certificate of clinical competency in Speech and Language Pathology (CCC-SLP).

January 1st, 2020

The IEEE Signal Processing Society, Bangalore Chapter and

Indian Institute of Science

Cordially invite you to the following talk on

New Twists for New Tricks, Making Audio Deep Learning Practical

Speaker: Prof. Paris Smaragdis, University of Illinois at Urbana-Champaign.

Date and Time: 6th Jan, 2020, 4pm, Refreshments: 3:45pm

Venue: ECE Golden Jubilee Seminar Hall.

Talk flyer

December 18th, 2019

IEEE Signal Processing Society, Bangalore Chapter and Department of EE, Indian Institute of Science Invite you to the following talk:

SPEAKER   :  Prof. Ardhendu Behera, Associate Professor (Reader), Edge Hill University, UK

TITLE          : “Computer Vision and Deep Learning – A Marriage of Neuroscience and Machine Learning”

Venue            :  MMCR Room No C241, First Floor, EE Dept.

Date & Time :  Dec 20, 2019, 04:00 PM


For almost 10 decades, human vision researchers have been studying how the human vision system has evolved. While computer vision is a much younger discipline, it has achieved impressive results in many detection and classification tasks (e.g. object recognition, scene classification, face recognition, etc.) within a short span of time. Computer vision is one of the fastest growing fields and one of the reasons is due to the amount of video/image data from urban environment growing exponentially (e.g. 24/7 cameras, social media sources, smart city, etc.). The scale and diversity of these videos/images make it very difficult to extract reliable information to automate in a timely manner. Recently, Deep Convolutional Neural Networks (DCNNs) have shown impressive performance for solving visual recognition tasks when trained on large-scale datasets. However, such progresses face challenges when rolling into automation and production. These include enough data of good quality, executives’ expectations about model performance, responsibility and trustworthiness in decision making, data ingest, storage, security and overall infrastructure, as well as understanding how machine learning differ from software engineering.

In this talk, I will focus on recent progress in advancing human action/activity and behaviour recognition from images/videos, addressing the research challenges of relational learning, deep learning, human pose, human-objects interactions and transfer learning. I will then briefly describe some of our recent efforts to adopt these challenges in automation and robotics, in particular human-robot social interaction, in-vehicle activity monitoring and smart factories.

Speaker Bio:

Ardhendu Behera is a Senior Lecturer (Associate Professor) in the Department of Computer Science in Edge Hill University (EHU). Prior to this, he held post-doc positions at the universities of Fribourg (2006-07) and Leeds (2007-14). He holds a PhD from University of Fribourg, MEng from Indian Institute of Science, Bangalore and BEng from NIT Allahabad. He is leading the visualisation theme of the Data and Complex Systems Research Centre at the EHU. He is also a member of Visual Computing Lab. His main research interests are computer vision, deep learning, pattern recognition, robotics and artificial intelligence. He applies this interest to interdisciplinary research areas such as monitoring and recognising suspicious behaviour, human-robot social interactions, autonomous vehicles, monitoring driving behaviour, healthcare and patient monitoring, and smart environments. Dr Behera has been involved in various outreach activities and some of his research are covered by media, press, newspaper and television.




December 17th, 2019

IEEE Signal Processing Society, Bangalore Chapter and Department of Computational and Data Sciences, 

Indian Institute of Science Invite you to the following talk:

SPEAKER   :  Prof. Dima Damen, Associate Professor (Reader), University of Bristol, UK 

TITLE          : A fine-grained perspective onto object interactions 

Venue            :  #102, CDS Seminar Hall

Date & Time :  Dec 26, 2019, 04:00 PM


This talk aims to argue for a fine-grained perspective onto human-object interactions, from video sequences. The talk will present approaches for determining skill or expertise from video sequences [CVPR 2019], assessing action ‘completion’ – i.e. when an interaction is attempted but not completed [BMVC 2018], dual-domain and dual-time learning [CVPR 2019, ICCVW 2019] as well as multi-modal approaches using vision, audio and language [ICCV 2019, BMVC 2019].

This talk will also introduce EPIC-KITCHENS [ECCV 2018], the recently released largest dataset of object interactions in people’s homes, recorded using wearable cameras. The dataset includes 11.5M frames fully annotated with objects and actions, based on unique annotations from the participants narrating their own videos, thus reflecting true intention. Three open challenges are now available on object detection, action recognition and action anticipation [


Dima Damen is a Reader (Associate Professor) in Computer Vision at the University of Bristol, United Kingdom. Received her PhD from the University of Leeds, UK (2009). Dima’s research interests are in the automatic understanding of object interactions, actions and activities using static and wearable visual (and depth) sensors. Dima co-chaired BMVC 2013, is area chair for BMVC (2014-2018), associate editor of IEEE TPAMI (2019-) and associate editor of Pattern Recognition (2017-). She was selected as a Nokia Research collaborator in 2016, and as an Outstanding Reviewer in ICCV17, CVPR13 and CVPR12. She currently supervises 6 PhD students, and 4 postdoctoral researchers. More details at: []

Host Faculty: Prof. Venkatesh Babu


                        ALL ARE WELCOME