Archive for the ‘Uncategorized’ Category

[Seminar] CDS : 25th April 2018 : “Powering fashion e-commerce with computer vision and deep learning “

Friday, April 20th, 2018

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

Invite you to the following talk:

Title: Powering fashion e-commerce with computer vision and deep learning

by Vishnu Vardhan Makkapati, Architect, Myntra,

Time & Date: 11:00 AM, Wednesday, April 25, 2018
Venue: CDS Seminar Hall (Room No: 102), IISc.


Fashion is a highly visual field. Images, though a rich source of domain information, are extremely subjective, and their interpretation is more art than science. Our goal is to teach a machine to interpret these varied images in a consistent manner, while eliminating subjectivity from the process. We have used our industry leading fashion catalog to understand and interpret the inherent fine-grained details in images. These details will help power very interesting use cases in fashion e-commerce such as cataloging, purchasing, personalisation etc. In this talk, we will present an overview of our work on mining catalog images using deep learning and computer vision. We will also discuss some of our recent work on generation of fashion designs using Generative Adversarial Networks.

Speaker Bio:
Vishnu Vardhan Makkapati received the B.E. (Honors) degree in electrical and electronics and the M.Sc. (Honors) degree in mathematics from the Birla Institute of Technology and Science, Pilani, India, in 2000, and the M.Sc. (Engg.) degree from the Indian Institute of Science, Bangalore, in 2007. He was with the IBM India Software Laboratory until April 2001 and, then, with the Honeywell Technology Solutions Laboratory, India, until December 2006, reaching the level of Principal Engineer. He was a Senior Scientist with Philips Research India until July 2015, where he most recently led the efforts on camera based vital signs. He is currently an Architect with Myntra. He holds six US patents with many others pending. He is a senior member of the IEEE.


[Seminar] ECE: 26th April 2018 : “Hearing Assessment Technologies (Audiometry) and some new challenges”

Friday, April 20th, 2018

Dept of ECE, Indian Institute of Science, Bangalore
IEEE Signal Processing Society, Bangalore Chapter

have arranged the following technical lecture; All are welcome.

Speaker: Dr. Madhuri Gore

Title: “Hearing Assessment Technologies (Audiometry)
and some new challenges”
Date/Time: 26th April 2018, Thursday, 4-5pm
Venue: Golden Jubilee Seminar Hall, ECE Dept, IISc
Tea/Coffee: 3.45pm

The auditory system has unique abilities that enable to person to carry out tasks such as listening in noise, detecting differences in sound object, localizing sounds, music appreciation and learning languages, activities that are deceptively simple. Hearing assessment (common parlance referred to as audiometry) refers to behavioural and physiological tests to assess function and sometimes “structure” of the different components of the auditory pathway. For this, a number of stimuli varying from the simple pure tones to synthetic speech which are presented in different modes, at different SNRs, in an attempt to simulate real life conditions, but more often to diagnose the problem. Creating appropriate stimuli and calibration is an issue.

Speaker Bio:
Dr. Madhuri Gore is Principal and Deputy Director (T) of “Dr.S.R.Chandrasekhar Institute of Speech and Hearing,” Bangalore. She has been in academics since 1993. She has vast clinical experience in hearing disorders since 1982. Her special clinical interest in Cochlear Implants, Auditory Plasticity, Auditory Neuropathy, Spectrum Disorder; also interested in Diagnostic Audiology, Psychophysics and Speech Perception.

[Seminar] CDS : 19th April 2018 : “Image-based Crowd Analytics”

Thursday, April 12th, 2018

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:

Title: Image-based Crowd Analytics
by Vishwanath A. Sindagi, Rutgers University,

Time & Date: 11:00 AM, Thursday, April 19, 2018
Venue: CDS Seminar Hall (Room No: 102), IISc.


The study of human behavior based on computer vision techniques has gained a lot of interest in recent years. In particular, the behavioral analysis of crowded scenes is of great interest due to a variety of reasons. Exponential growth in the world population and the resulting urbanization has led to an increased number of activities involving high density crowd such as sporting events, political rallies, public demonstrations, thereby resulting in more frequent crowd gatherings in the recent years. In such scenarios, it is essential to analyze crowd behavior for better management, intelligence gathering, safety and security. In this talk, I will present some of my recent work on developing algorithms for crowd analytics, including crowd counting from unconstrained imagery, crowd segmentation and human detection from crowded scenes. I will conclude my talk by describing several promising directions for future research.


Vishwanath is a second year PhD student in Dept. Of Electrical & Computer Engineering at Rutgers University. He is being advised by Prof. Vishal M Patel. Prior to joining Rutgers, he worked for Samsung R&D Bangalore and AllGo Embedded Systems. He graduated from IIIT-Bangalore with a Master’s degree in Information Technology. His current research is on computer vision and machine learning with a specific focus on crowd analytics, face detection, applications of generative modeling, domain adaptation and low-level vision.


Apr. 3, 2018: : “Towards Phonologically Motivated Sign Language Processing”

Friday, March 30th, 2018

The IEEE Signal Processing Society Bangalore Chapter and
Department of Electrical Engineering, Indian Institute of Science, Bangalore

invite you to a lecture by

Dr. Mathew Magimai Doss from Idiap, Martigny and EPFL, Switzerland

on the following topic:

“Towards Phonologically Motivated Sign Language Processing”

Sign language is a mode of communication commonly employed by the Deaf community to communicate with each other as well as to communicate with the Hearing community. SMILE is a Swiss NSF funded Sinergia project involving sign language technologists and sign linguists that aims to develop a sign language learning system for Swiss German sign language (DSGS). In this talk, I will present recent developments in the SMILE project. More specifically, I will present (a) development of SMILE DSGS dataset and (b) development of a phonologically motivated sign language assessment approach using hidden Markov models and artificial neural networks, akin to articulatory feature-based speech processing.

Venue: Multimedia Classroom, Department of Electrical Engineering, IISc

Time: 4 PM, April 3, 2018 (Coffee will be served at 3.45 PM)

Speaker Biography:
Dr. Mathew Magimai Doss received the Bachelor of Engineering (B.E.) in Instrumentation and Control Engineering from the University of Madras, India in 1996; the Master of Science (M.S.) by Research in Computer Science and Engineering from the Indian Institute of Technology, Madras, India in 1999; the PreDoctoral diploma and the Docteur dès Sciences (Ph.D.) from Ecole Polytechnique Fédérale de Lausanne (EPFL), Switzerland in 2000 and 2005, respectively. He was a postdoctoral fellow at International Computer Science Institute (ICSI), Berkeley, the USA from April 2006 until March 2007. Since April 2007, he has been working as a permanent researcher in the Speech and Audio Processing group at Idiap Research Institute, Martigny, Switzerland. He is also a lecturer at EPFL. He is a senior area editor of the IEEE Signal Processing Letters. His main research interest lies in signal processing, statistical pattern recognition, artificial neural networks and computational linguistics with applications to speech and audio processing and multimodal signal processing.

IEEE SP Chapter talk : 22nd February, 2018 : “Deep Representations, Adversarial Learning and Domain Adaptation for Some Computer Vision Problems”

Sunday, February 18th, 2018

IEEE Signal Processing Society, Bangalore Chapter, IEEE Bangalore Section,
Department of Electrical Engineering, Indian Institute of Science

invite you to a seminar by

Prof. Rama Chellappa
Distinguished University Professor,
University of Maryland, College Park

Date: 22 February, 2018, Thursday,
Time: 4:00 pm (coffee will be served at 3:45pm)

Venue: Multimedia Classroom, Department of Electrical Engineering, IISc

Title: Deep Representations, Adversarial Learning and Domain Adaptation for Some Computer Vision Problems

Recent developments in deep representation-based methods for many computer vision problems have knocked down many research themes pursued over the last four decades. In this talk, I will discuss methods based on deep representations, adversarial learning and domain adaptation for designing robust computer vision systems with applications in unconstrained face and action verification and recognition, expression recognition, subject clustering and attribute extraction. The face and action recognition system being built at UMD is based on fusing multiple deep convolutional neural networks (DCNNs) trained using publicly available still
and video face data sets. I will then discuss some new results on generative adversarial learning and domain adaptation for improving the robustness of the recognition system.

Biography: Prof. Rama Chellappa is a Distinguished University Professor, a Minta Martin Professor of Engineering and Chair of the ECE department at the University of Maryland. His current research interests span many areas in image processing, computer vision, machine learning and pattern recognition. Prof. Chellappa is a recipient of an NSF Presidential Young Investigator Award and four IBM Faculty Development Awards. He received the K.S. Fu Prize from the International Association of Pattern Recognition (IAPR). He is a recipient of the Society, Technical Achievement and Meritorious Service Awards from the IEEE Signal Processing Society. He also received the Technical Achievement and Meritorious Service Awards from the IEEE Computer Society. Recently, he received the inaugural Leadership Award from the IEEE Biometrics Council. At UMD, he received college and university level recognitions for research, teaching, innovation and mentoring of undergraduate students. In 2010, he was recognized as an Outstanding ECE by Purdue University. He received the Distinguished Alumni Award from the Indian Institute of Science in 2016. Prof. Chellappa served as the Editor-in-Chief of PAMI. He is a Golden Core Member of the IEEE Computer Society, served as a Distinguished Lecturer of the IEEE Signal Processing Society and as the President of IEEE Biometrics Council. He is a Fellow of IEEE, IAPR, OSA, AAAS, ACM and AAAI and holds six patents.

IEEE Signal Processing Society,
Bangalore Chapter

Seminar [Jan 9, 2018]: Audio and Acoustics Signal Processing: the Quest for High Fidelity Continues

Thursday, January 4th, 2018

The IEEE Signal Processing Society Bangalore Chapter

invite you to a talk on

Audio and Acoustics Signal Processing: the Quest for High Fidelity Continues

Prof. Dr.-Ing. Dr. rer. nat. h.c. mult. Karlheinz Brandenburg
Head of the electronic media technolgy lab
Director of the Fraunhofer Insitute for Digital Media Technology (IDMT)

Date and time: January 9, 2018; 9:30 AM

Venue: EC 1.06, Department of Electrical Communication Engineering, Indian Institute of Science (IISc).

The dream of high fidelity continues since more than 100 years. In the last decades, signal processing has contributed many new solutions and a vast amount of additional knowledge to this field. This includes mp3 and other audio coding algorithms which changed the way we listen to music.

Current research includes the following areas:
– Music Information Retrieval (MIR), helping us to to find music or even learn playing instruments
– Immersive technologies to really get the best illusion into movie theaters and eventually our homes
– Learning more about hearing, how the ear and the brain work together when we listen to sounds
The talk will explain the basics and touch on some current research topics in these fields.

Biography of the speaker:
Karlheinz Brandenburg is the inventor of MP3 (

[Seminar]: ECE: 21 Dec. 2017: “Beyond Worst Case – When data size increases faster than Moore’s law, … “

Monday, December 18th, 2017

Dear All:

IEEE Signal Processing Society, Bangalore Chapter, IEEE Bangalore Section,
and Dept. of ECE, IISc cordially invite you to the following talk:

Title: Beyond Worst Case – When data size increases faster than Moore’s law, one has to sacrifice worst-case performance to improve the average-case performance.
Speaker: Vivek Bagaria, Stanford University, USA
Time and Date: 4pm, Thursday, Dec. 21, 2017
Venue: Golden Jubilee Hall, ECE Department, IISc Bangalore


High dimensional median: Computing the medoid of a large number of points in high-dimensional space is an increasingly common operation in many data science problems. We present an algorithm Med-dit which computes the medoid (with high probability) in O(n log n) steps. Med-dit is based on a connection with the multi-armed bandit problem. We empirically evaluate the performance of Med-dit on the Netflix-prize and the single-cell RNA datasets, containing hundreds of thousands of points living in tens of thousands of dimensions, and observe a 5-10x improvement in performance over the current state of the art.

Planted Traveling salesman problem (TSP): Consider a complete graph K_n. The edge weights of a planted hamiltonian path are drawn from distribution P and all other edge weights are drawn from distribution Q. We present an almost-linear time algorithm which solves planted-TSP above information theoretic limit (for certain class of distributions P and Q such as gaussian, poisson etc). We apply this algorithm on Chicago-seq dataset and show a potential improvement in the quality of de novo dna assembly.


Vivek Bagaria is a doctoral student at Stanford University, advised by Prof. David Tse. His google scholar profile can be found here:

[Seminar] CDS : 22nd December 2017 : “Connecting vision and language for Interpretation, Grounding and Imagination”

Saturday, December 16th, 2017

Dear All,

IEEE Signal Processing Society, Bangalore Chapter, IEEE Bangalore Section,
and Department of Computational and Data Sciences, Indian Institute of Science

Invite you to the following talk:

Title: Connecting vision and language for Interpretation, Grounding and Imagination
by Ramakrishna Vedantam, Georgia Tech,

Time & Date: 4:00 pm, Friday, December 22, 2017
Venue: CDS Seminar Hall (Room No: 102), IISc.

The goal of this talk is to explore how modeling deep interactions between vision and natural language can derive more humanlike inferences from machine learning models. We will consider various situations where humans are able to make intuitive inferences, but where machines are not, and show how appropriate algorithmic or computational choices can improve the inferences made by machines and make them more humanlike.

In pursuit of this overarching goal, I will look at three focus areas: interpretation, grounding, and visual imagination. In interpretation, I will study how to go from computer vision to natural language. Specifically, the focus will be on image captioning: the problem of describing an image with a natural language description. Here, I study how to formalize the task of image captioning and create evaluation schemes to make progress towards generating more `human-like’ descriptions and generate image captions which are more aware of the context in which we want to describe an image.

In the second part, I will consider the inverse problem of “grounding” — associating symbols such as “car” with what they refer to in the physical world. A key focus will be on learning word representations grounded in vision or sound, and a study of how such representations can lead to improved retrieval and commonsense reasoning.

Finally, I will talk about recent work on the problem of imagination — consider how easy it is for people to imagine what a “purple hippo” would look like, even though they do not exist. If we instead said “purple hippo with wings”, they
could just as easily create a different internal mental representation, to represent this more specific concept. To assess whether the person has correctly understood the concept, we can ask them to draw a few sketches,
to illustrate their thoughts. We call the ability to map text descriptions of concepts to latent representations and then to images visually grounded semantic imagination. My work will explore how one can modify a class of joint latent variational autoencoder models to perform such grounded semantic imagination.


Ramakrishna Vedantam is a PhD. student in the school of Interactive Computing at the College of Computing at Georgia Tech since 2017. His advisor is Devi Parikh. Before moving to GT, he was a PhD. student at the Bradley Department of ECE at Virginia Tech.

During his Ph.D. he has spent time at Facebook AI Research (Summer, 2017), Google Research (Winter, 2017, and Summer,2016) and INRIA-Saclay (Summer, 2014). He received his M.S. degree in ECE from Virginia Tech in 2016, and his bachelors in ECE from the International Institute of Information Technology (IIIT) – Hyderabad in 2013.

He is interested in problems at the intersection of vision and language, as well as generative models and variational
inference. In the past he has worked on topics like evaluation and generation of image captions, grounding language into visual cues, etc. He was the recipient of the Outstanding Reviewer Award at CVPR, 2017, and a finalist for the Adobe Research Fellowship in 2016.


Talk [Dec 20, 2017] – Positive trigonometric polynomials: Application to spectral super-resolution

Saturday, December 16th, 2017

The Department of Electrical Engineering and
The IEEE Signal Processing Society Bangalore Chapter

invite you to a talk on

“Positive trigonometric polynomials: Application to spectral super-resolution”

Dr. Kumar Vijay Mishra, IIHR – Hydroscience & Engineering, The University of Iowa, USA.

Date and time: December 20, 2017; 4.30 PM

Venue: Multimedia Classroom, Department of Electrical Engineering, IISc.

We address the problem of super-resolution frequency recovery using prior knowledge of the
structure of a spectrally sparse, undersampled signal with frequencies lying anywhere in the
continuous domain [0, 1]. We devise a general semidefinite program (SDP) to recover these
frequencies using theories of positive trigonometric polynomials (PTP). Our theoretical analysis
shows that given sufficient prior information, perfect signal reconstruction is possible using signal
samples no more than thrice the number of signal frequencies. We extend our PTP formulations to
solve an open problem on the formulation of an equivalent positive semidefinite program for atomic
norm minimization in recovering signals with d-dimensional (d greater than or equal to 2) off-thegrid
frequencies. Finally, we combine SDP and l1-minimization to develop fast versions of our

Biography of the speaker:
Dr. Mishra obtained Ph.D. in electrical engineering and M.S. in mathematics from The University of
Iowa in 2015 and M.S. in electrical engineering from Colorado State University in 2012 while
working on NASA GPM-GV mission weather radars. He received B. Tech. summa cum laude
(Hons., Gold Medal) in electronics and communication engineering from the National Institute of
Technology, Hamirpur in 2003. During 2003-2007, he worked as a research scientist at eLectronics
and Radar Development Establishment (LRDE), Defence Research and Development
Organization, Bengaluru on military surveillance radars. He was a research intern at Mitsubishi
Electric Research Labs (Cambridge) and Qualcomm (San Jose) in 2015. During 2015-2017, he
was Andrew and Erna Finci Viterbi and Lady Davis postdoctoral fellow at the Faculty of Electrical
Engineering, Technion – Israel Institute of Technology. He is the recipient of Royal Meteorological
Society Quarterly Journal Editor’s Prize (2017), Lady Davis Fellowship (2016-17), Andrew and
Erna Finci Viterbi Fellowship (twice awarded, 2015, 2016), Technion EE Excellent Undergraduate
Adviser Award (2017), DRDO LRDE Scientist of the Year Award (2006), Cornell Base-of-Pyramid
Narrative Competition Prize (2009), Altera Forum Guru Challenge Winner (2008), and NITH Best
Student Award (2003). His research interests include radar systems-theory-and-hardware, signal
processing, radar polarimetry, and electromagnetics. His recent research is focused on radar
applications in sub-Nyquist processing, spectrum sharing, autonomous vehicles, MIMO, cognitive
radars and deep learning.

IEEE Signal Processing Society
Bangalore chapter

Talk by Dr. Arun Kumar on 4 Dec, 2017 at 3 PM TITLE OF THE TALK: ” Acoustic Vector Sensors and related Signal Processing for Air and Underwater Applications”

Wednesday, November 22nd, 2017

The IEEE Signal Processing Society, Bangalore Chapter and

The Department of Electrical Communication Engineering, IISc

invite you to the following lecture.

Name of the speaker: Dr. Arun Kumar

Title of the Talk: Acoustic Vector Sensors and related Signal Processing for Air and Underwater Applications

Date and Time: 4th Dec, 2017 at 3PM

Venue: Golden Jubilee Seminar Hall of ECE Department


Acoustic emissions from radiating sources or targets, in either air or underwater medium, can be used to detect, localize, and track them passively. The classical method of estimating the Direction of Arrival (DOA) of a radiating acoustic source is to use a spatially distributed array of pressure sensors i.e. hydrophones or microphones. The DOA can also be estimated by collocated measurement of the particle velocity and the pressure of the signal that gives the acoustic intensity vector. An Acoustic Vector Sensor (AVS) consists of three orthogonally oriented velocity sensors and a pressure hydrophone or microphone, all spatially collocated in a point-like geometry. The collocated
measurement viz. the particle velocity that is a vector quantity along with the pressure results in a 4-dimensional vector that is recorded as function of time. This measurement can give improved DOA estimation even with a single AVS or smaller AVS arrays as compared with only pressure sensor arrays requiring comparatively large apertures. By making additional use of the velocity information, systems built around AVS have numerous advantages over pressure sensor arrays.

In the talk, the use of AVS will be motivated, followed by a presentation of the salient features of an AVS, and two methods for measuring the particle velocity. The practical issues related to the two methods of constructing AVS will then be discussed. The work done by our research group at IIT Delhi for the design, development and evaluation of air and underwater acoustic vector sensors and their related signal processing algorithms will also be presented.

Short Biography of Prof. Arun Kumar

Arun Kumar received the B.Tech, M.Tech, and PhD degrees in Electrical Engineering from IIT Kanpur. He was a Visiting Researcher at the University of California, Santa Barbara, USA, for 2 years. Since 1997, he has been with the Centre for Applied Research in Electronics, IIT Delhi, where he is Professor. He was Head of the Centre for 4 years. He also served as Head of Instrument Design and Development Centre of IIT Delhi. His research interests are in digital signal processing, human and machine speech communication technologies, underwater and air acoustics, acoustic imaging, acoustic vector sensors, and multi-sensor data fusion for mobile and wearable devices.

Arun Kumar is an inventor on 8 US patent applications (2 granted and 6 pending). He has published over 100 papers in refereed journals and conferences, and has supervised 55 funded R&D projects. These have led to 20 Technology and know-how transfers. Several of these technologies such as High Speed VLF Communication Modem for communication with submerged submarines are deployed in the field and in practical use.

He has served on several National level committees. He is Editor of the IETE Journal of Research. He is also co-founder and non-executive Director of two technology companies working in the areas of speech technologies and DSP products.