Amit Kale

Web: http://www.cvisionresearch.net Email: [email protected]

Objective To obtain a challenging senior leadership position in the areas of AI and ML.

Education Ph.D., Electrical and Computer Engineering, October 2003; GPA: 3.8
“Algorithms for Gait-Based Human Identification from a Monocular Video Sequence”
University of Maryland at College Park, College Park, MD

M.Tech., Systems and Control Engineering, Spring 1998; GPA: 3.95

Indian Institute of Technology, Bombay, Mumbai India.

B.E., Electrical Engineering, Fall 1996; GPA: 3.75

Victoria Jubilee Technical Institute (affiliated to University of Bombay), Mumbai, India.

Work Experience Chief Scientist

Robert Bosch Research and Technology Center, Bangalore India
October 2022- date
As a chief scientist in the area of data management for AI, I focus on problems related to enabling
the data loop for AI. This includes supervising technology development to address every phase
of the data loop starting with acquisition and ingest, data selection, labeling and QC, data
re-use and usage of synthetic data for model development, approaches for validation and release
of AI algorithms. In this endeavor I lead a global team of 44 research scientists focussing on
some of these aspects. I have had high impact contributions in AI based smart data curation,
data-reuse and assisted annotation for data management (¿15 mio Euro bottom line impact)
and 20+ Invention Reports filed in the AI domain. This role is in addition to my director role.

Director of Research
Robert Bosch Research and Technology Center, Bangalore India
April 2022- date

In this role, I am responsible for the technology strategy, project acquisition, leadership and
further development of the department of corporate research in India. I lead a group of 30
scientists in areas of AI, CV, ML and numerical methods for embedded software.

Principal Senior Expert and Group Manager

Computer Vision and Machine Learning
Robert Bosch Research and Technology Center, Bangalore India
June 2016- March 2022 In this role, I am responsible for the technology strategy, project acquisi-
tion, leadership and further development ofthe research group for computer vision and machine
learning and in India and its projects. My primary focus is on developing core technologies
that improve diffrent aspects of the data loop for AI development. This includes the phases
of acquisition and ingest, smart data selection and curation of data for labeling, search and
retrieval of data, identifying problematic images during validation and testing. We make use of
advanced deep learning approaches to achieve these goals in multiple domains such autonomous
driving, security and surveillance, retail, industrial imaging etc. We have built a number of high
value scalable assets with relevant IP which we have transferred to BUs both globally and in
the region for various applications.

Head of Research Group

Imaging and Computer Vision
Siemens Corporate Technology India
July 2012- to June 2016

In this role, I was responsible for technology strategy, project acquisition and the exploration
of research projects for the imaging and computer vision program in India. My work involved
exploring the clinical collaboration eco-system in India. Our work was demonstrated and pub-
lished in leading conferences and journals like RSNA and Clinical Radiology Journal. My task
involved interfacing the imaging and computer vision technology field and developing meaning-
ful technology solutions to address needs of clinicians in India. I also worked jointly with the
teams in the Development Center to provide enabling technologies towards building a proof of
concept of ideas from team members working on specific applications. During my tenure here we
developed successful partnerships with different business units in Siemens worldwide including
our MR, CT, AX and Computer-aided diagnostics delivering high quality prototypes, while also
contributing to the IP portfolio of Siemens, globally. Our work has resulted in 10 granted US
patents with many more in the pipeline. I was also responsible for recruitment, mentoring and
career development of a team of scientists and engineers in the research group 10 headcounts.
I have helped develop the business case for novel applications of MR imaging for preventive
healthcare in India and helped do a gap and need analysis for our x-ray product portfolio as
well. I manage the budget and financing for my team. The total volume of funding I managed
was close to Euro 500k.

Program Manager
Imaging and Intelligent Signal Processing
Siemens Corporate Technology India
Feb 2007 – June 2012

In this role I was responsible for leading and driving the mission of the imaging program in
India. I was responsible for developing technical expertise by mentoring team members in
computer vision applications in medical imaging, video surveillance and other domains. Other
responsibilities included staffing the program as per target, financial budgeting, driving univer-
sity relationships, scouting opportunities in Indian markets for novel imaging solutions, creating
IP for the company and publications in reputed conferences and journals. Some of the top-
ics I worked on which formed the basis for several BU funded research proposals and patents
included:
• Automated Analysis of ECG and IECG signals
• Development of a robust low cost patient positioning and monitoring system for stereo-
tactic radiotherapy of brain tumors.
• Robust Vessel segmentation in Angio Images for accurate 2D/3D registration within car-
diac total occlusion project
• Human Activity Analysis
• Robust Face Detection and Tracking

Postdoctoral Research Associate and Assistant Research Professor

Department of Computer Science and
Center for Visualization and Virtual Environments
University of Kentucky Lexington, KY
Oct 2003 – January 2007

In this role I was responsible for development of novel applications in augmented environments
using computer vision techniques. I supervised research conducted by Masters students and
interns in the following areas:
• Augmented Reality systems using projector and camera including for Camera Projector
based Scene Augmentation and Human Computer Interaction in Projector Camera Sys-
tems
• Semi-Supervised Learning for Background Subtraction and Object Detection
• Robust Visual tracking using parametric illumination models

Research Assistant
Center for Automation Research
 University of Maryland, College Park
College Park, MD
Aug 1998 – Oct. 2003 In the course of my Ph.D. I developed robust algorithms for human
identification using gait. I worked on template matching and hidden Markov model based
models for gait recognition leading to a top performing system for gait recognition. I also
worked on approaches for view invariant gait recognition and fusion of gait and face for robust
human identification. My work on this topic is widely cited.

Honors • Best Paper award at the IEEE Conference on Parallel Distributed and Grid Computing
2012 for our paper ”Fast 3D Structure Localization in Medical Volumes using CUDA-
enabled GPUs” co authored with Sharan Vaswani, Naga Vydyanathan and Rahul Thota
• Best paper award (poster) at British Machine Vision Conference (BMVC) 2005 for our
paper entitled “”Offline Generation of High Quality Background Subtraction Data” co
authored with Etienne Grossmann, Christopher Jaynes and Samson Cheung.
• My work in human motion analysis and tracking has been recognized to be among the top
percentile of the accepted papers in the conferences pertinent to my field.

Professional • Jury member for grant proposal reviews at Grand Challenges Karntataka for Agriculture,
Activities Traffic management solutions 2018, 2019, 2020
• Program Co-chair for IEEE Workshop on Multimodal Surveillance - Sensors, Algorithms
and Systems 2007
• Reviewer for IEEE Transactions on Image Processing, IEEE Transactions on Pattern Anal-
ysis and Machine Intelligence, IEEE Transactions on Information Forensics and Security,
IEEE Transactions on Systems Man and Cybernetics, ACM Transactions on Multimedia
Computation, EURASIP Journal on Signal processing, IEEE Transactions on Biomedical
Engineering.
• Program Committee member for IEEE CVPR 2006, IEEE ICIP 2006-2020 , ECCV 2008-
2010, ICCV 2009, ACCV 2010, ACCV 2012, ICVGIP 2012, ACM Multimedia 2004, Work-
shop on Applications of Computer Vision 2005, International Conference on Circuits and
Systems 2005, Workshop on Real time Vision for Human Computer Interaction 2004.

Featured • Research profile at Bosch Research https://www.bosch.com/research/know-how/research-

Activities experts/amit-kale/
• Research Blog article on Improving annotation efficiency for deep learning algorithms
for computer vision https://www.bosch.com/stories/improving-annotation-efficiency-for-
deep-learning-algorithms-for-computer-vision/

Invention 2020/6494 System and Method for Object Search in Images 1. Kale Amit Arvind(CR/RTC2-
Reports Filed IN) 2. Singh Sonam(CR/RTC2-IN) 3. Jitendra Yasaswi Bharadwaj Katta(CR/RTC2-IN) 4.
from Bosch Mullick Koustav(CR/RTC2-IN) 394133 22.12.2020

2020/6495 System and Method for Object Search in Images using Incremental Relevance Feed-
back 1. Kale Amit Arvind(CR/RTC2-IN) 2. Singh Sonam(CR/RTC2-IN) 397231 19.07.2021

2020/6497 System and Method for Object Search in Images using Augmentation 1. Kale Amit
Arvind(CR/RTC2-IN) 2. Singh Sonam(CR/RTC2-IN) 394135 19.07.2021

2020/6498 Automatic Query Expansion for Searching an Object in Images using Relevance Feed-
back 1. Kale Amit Arvind(CR/RTC2-IN) 2. Singh Sonam(CR/RTC2-IN) 394134 22.12.2020

2020/7905 System and Method for Domain adaptation using intelligent pairing of data and
novel cost function to enforce semantic similarity 1. Mullick Koustav(CR/RTC2-IN) 2. Jain
Harshil(CR/RTC2-IN) 3. Kale Amit Arvind(CR/RTC2-IN) 397232 30.07.2021
 2021/0098 System and Method for Image Content Translation across different Cameras 1. Mul-
lick Koustav(CR/RTC2-IN) 2. Jain Harshil(CR/RTC2-IN) 3. Kale Amit Arvind(CR/RTC2-IN)
397244 30.07.2021

2021/0440 System and Method for Unsupervised Selection of Images for Labeling 1. Kale
Amit Arvind(CR/RTC2-IN) 2. Jitendra Yasaswi Bharadwaj Katta(CR/RTC2-IN) 3. Mullick
Koustav(CR/RTC2-IN) 398246 26.08.2021

2021/5342 System and Interactive User Interface for Search and Curation of Object Attributes in
Images 1. Jitendra Yasaswi Bharadwaj Katta(CR/RTC2-IN) 2. Kale Amit Arvind(CR/RTC2-
IN) 3. Raju C Devaraja Iyengar(BGSW/ERD1) 4. Akash Rajendranath Kotian(BGSW/ERD1)
399135 29.10.2021

2021/5878 System and Method to find Corner Cases using Class specific, Unsupervised, Deep
Feature based Active-learning. 1. Mullick Koustav(CR/RTC2-IN) 2. Kale Amit Arvind(CR/RTC2-
IN) 3. Jitendra Yasaswi Bharadwaj Katta(CR/RTC2-IN) 4. Shanbhag Mahendra(BGSW/ERD1)
399131 29.11.2021

2021/5927 System and Interactive User Interface for Quality Check of labelled data 1. Kumar
Abhijeet(CR/RTC2-IN) 2. Mullick Koustav(CR/RTC2-IN) 3. Kale Amit Arvind(CR/RTC2-
IN) 4. Vinoj John Hosan Ebenezer Koil Pillai(BGSW/ERD1) 399132 30.12.2021

2021/6389 System and Method for computer vision based vehicle speed detection 1. Chiranjeevi
Pojala(CR/RTC2-IN) 2. Kale Amit Arvind(CR/RTC2-IN)

2021/8567 System and Method for Visual Templates based Acquisition of Images 1. Singh
Sonam(CR/RTC2-IN) 2. Kale Amit Arvind(CR/RTC2-IN)

2021/8568 System and Method for Acquisition of Images using Reweighted Features 1. Singh
Sonam(CR/RTC2-IN) 2. Kale Amit Arvind(CR/RTC2-IN)

2021/8569 : System and Method for Mining Similar Failures of a deep learning model using
Human in the Loop 1. Singh Sonam(CR/RTC2-IN) 2. Kale Amit Arvind(CR/RTC2-IN)

2021/8570 System and Method for Iterative Refinement and Curation of Images driven by Visual
Templates 1. Singh Sonam(CR/RTC2-IN) 2. Kale Amit Arvind(CR/RTC2-IN)

2022/1046 System for Predicting the Labelling Complexity of an Image 1. Humbarwadi Sri-
hari(BGSW/ERD1) 2. Agrawal Ashutosh Kumar(BGSW/ERD1) 3. Kale Amit Arvind(CR/RTC2-
IN)

Patents ”A fast and robust Technique for Segmenting Angiogram Images”, US Patent 8077954 granted
13.12.2011

”A robust and fast approach to automated coronary artery segmentation using cardiac phase
based tracking and a novel vessel detection criterion”, Vipin Gupta, Amit Kale and Hari Sundar.
US patent 8422754 Granted 16.04.2013

”A robust and accurate approach to automatic Blood Vessel detection and segmentation from
Angiography X-ray images using multi-stage Random Forests”, Vipin Gupta, Amit Kale and
Hari Sundar. US patent 8488852. Granted 16.07.2013

”An Automatic Non parametric Approach for classification and segmentation of ECG signals
with high accuracy”, Kaustubh Kulkarni, Stefan Kimmer and Amit Kale. US Patent 8880352.
Granted 4.11.14

”A robust and accurate algorithm for automated measurement of intervals in an intracardiac

 eletrogram ” Venkata Suryanarayana and Amit Kale. US Patent 9226682. Granted 5.1.16

”Determining plaque deposits in blood vessels” Yogesh Bathina, Parmeet Singh Bhatia, Rajen-
dra Prasad Jadiyappa, Amit Kale, US Patent 9462987. Granted October 11, 2016.

”Automatic assessment of perceptual visual quality of different image sets” Parmeet Bhatia,
Amit Kale US Patent 14/752395. Granted April 23, 2018

”Automatic detection of medical image acquisition protocol”, Parmeet Bhatia, Amit Kale, US
Patent 20170185713. Granted June 1 2020.

”Method and system for personalizing a vessel stent” Y.Bathina, A. Piriyakumar, R. Jadiyappa,
Amit Kale, US Patent 10898266. Granted Jan 26, 2021.

Selected Journal “On Kharitonov’s theorem without invariant degree assumption ”Amit Kale and Andre’ L.Tits.
Publications and Technical Report.Institute for Systems Research.1999. Automatica 36-7 pp 1075-1076 July 2000.
Book Chapters
“Identification of Humans Using Gait” Amit Kale, A. N Rajagopalan, A. Sundaresan, N. Cun-
toor, A. RoyChowdhury, V Kruger, Rama Chellappa IEEE Transactions on Image Processing
September 2004

“Gait-based human identification using appearance matching” Amit Kale, Naresh Cuntoor, B
Yegnanarayana, A N Rajagopalan and Rama Chellappa ”Optical and Digital Techniques for
Information Security” Ed. B Javidi et.al Springer Verlag 2004.

“Particle filter with Mode Tracker for Visual tracking across Illumination Changes” Samarjit
Das, Amit Kale, Namrata Vaswani. ”IEEE Trans. on Image Processing. April 2012

“Non-invasive quantification of hepatic steatosis in living, related liver donors using dual-echo
Dixon imaging and single-voxel proton spectroscopy” S Krishan D Jain Y Bathina A Kale N
Saraf S Saigal N Choudhary S.S. Baijal A Soin, Journal of Clinical Radiology. Nov 2015.

Selected “Self-Supervised Learning for texture classification using limited labeled data”, Sahana Prabhu,
Refereed Jitendra Yasaswi and Amit Kale, ICIP 2022.
Conference
Proceedings “Volumetric texture modeling using dominant and discriminative binary patterns”, Parmeet S.
Bhatia, Amit Kale, Zhigang Peng SPIE Medical Imaging 2019

“Automatic Optic Disk and cup segmentation of fundus images using deep learning”, Venkata
Gopal Edupuganti, Akshay Chawla, Amit Kale. ICIP 2018

“Fast 3D Structure Localization in Medical Volumes using CUDA-enabled GPUs”,Proceedings

of the 2nd IEEE International Conference on Parallel Distributed and Grid Computing 2012,
Sharan Vaswani, Rahul Thota, Nagavijayalakshmi Vydyanathan and Amit Kale, Won the best
paper award at PDGC 2012

“A robust and accurate approach to automatic Blood Vessel detection and segmentation from
Angiography X-ray images using multi-stage Random Forests”, Vipin Gupta, Amit Kale and
Hari Sundar, Proc SPIE Medical Imaging 2012

“An Unsupervised Framework for Action Recognition Using Actemes”, Kaustubh Kulkarni,
Edmond Boyer, Radu Horaud, and Amit Kale, Proc. of ACCV 2010.

“A Framework for indexing human actions in video”,Kaustubh Kulkarni, Srikanth Cherla, Amit
Kale and V.Ramasubramanian. Proceedings of the 1st workshop on Machine Learning for Visual
Motion Analysis, 2008.
“A Joint Model of Illumination and Shape for Visual tracking” Amit Kale and Christopher
Jaynes. Proceedings of IEEE Computer Society CVPR 2006.

“Offline Generation of High Quality Background Subtraction Data” Etienne Grossmann, Amit
Kale, Christopher Jaynes and Samson Cheung Proceedings of British Machine Vision Confer-
ence, September 2005, Oxford Brookes,UK. (Best Poster Award at BMVC)

“Towards Interactive Generation of ”Ground Truth” in Background Subtraction from Partially

Labeled Examples” Etienne Grossmann, Amit Kale, Christopher Jaynes Proceedings of the
IEEE Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveil-
lance, October 2005 Beijing China.

“The Smart Bookshelf: A study of camera projector scene augmentation of an everyday environ-
ment ” Danny Crasto, Amit Kale and Christopher Jaynes Proceedings of the IEEE Workshop
on Applications of Computer Vision, Colorado Springs CO January 2005

“Epipolar Constrained User Push Button Selection in Projected Interfaces ” Amit Kale, Kenneth
Kwan and Christopher Jaynes Proceedings of the 1st IEEE workshop on Real time Vision for
Human Computer Interaction, Washington DC June 2004

RCSfile: resume-us.tex,v Date: 2003/04/24 03:55:41 Revision: 2.4