design webfront.PNG
 

Who am I ?

A lil bit about me

I am currently working as a research associate at Games 24x7 (company In the space of science of Gaming), India, focusing on developing technology to enhance user interaction using Cognitive Neuroscience and AI. Here, I have also deployed solutions in the real through some of the latest products designed in the organization. I collaborate with MIT Media lab (Camera Culture group) and Cambridge University ( Dr. Andreas Bender’s Lab).

With MIT Media Lab, I am working on developing a Data-Driven Deep Agent-Based Model and Performing Latent space interpolation using Vector Quantized Variational Autoencoder. At the University of Cambridge, I am currently exploring Graph Neural Networks to accelerate drug discovery. I have also held research positions at the Alan Turing Institute ( Got the perseverance award), Indian Institute of Technology – Madras, Ericsson Research, and Georgia State University ( NSF Funded project).

I am looking forward for a Ph.D. position to take a deep dive into a research problem to create a broad impact. Thanks for visiting my profile! Have a great day.

Scroll down to know "How Organized and Unorganised" I am!!!

sethulinkdp.png
 
sethulinkdp_edited.jpg

Education

Alma Matter

June 2016 -  June 2020

Vellore Institute of Technology

Graduated with a B.Tech in Electronics and Communication also finished with an Honors Degree which was only awarded to 11 students among 900 students. Recipient of National Merit Scholarship with an All India Rank of 986 among 300k Students. Finished my Bachelors as a University Special Achiever which is one of the highest honors awarded to a student - Batch 2016-2020 -  Please click here to see the award

Transcripts

April 2014 - April 2016

Kendriya Vidyalaya, Island Grounds

Passed out as a National Rank Holder with 99% in Biology, CBSE Exams. Was an active member of the literature and Science Club where I won the Regional Essay writing competition and National Science Exhibition. Served as the School captain during my final year. 

Transcripts

April 2008 - April 2014

Kendriya Vidyalaya, Army Service Crops

Passed out with 98% in Secondary School Examination. Achieved national rank in Science. Was awarded a cash prize of 5000Rs by the Central Board of Secondary Education for the students who are among the top 0.1% in the nation. Won Multiple awards at the school level in science and arts and also won the South India's Skit competition

Transcripts
 

Professional Credentials

Roles & Responsibilities

playgamesupload_edited.jpg

Research Associate , Play Games 24x7

May 2021 - Present

  • Developed a creative recommendation system based on cognitive neuroscience to estimate the performance of advertisements which boosted the Click Through Rate by 0.31%.

  • Using the BOLD 5000 fMRI dataset, developed a model to estimate visual saliency and cognitive load for images and videos which is in the process of being Patented (US). (Prospective Submission to Nature)

  • Currently working on understanding affective gap caused by variations in personality along with creating a model to estimate the Valence and arousal for multimedia in the Wild. (Prospective Submission to ECCV 2022)

IITM 2.PNG

Project Associate, Indian Institute of Technology Madras

June 2020 - Present

  • Principal Investigator: Dr. Raghunathan Rengaswamy, Marti Mannariah Gurunath Institute Chair Professor, Dean of Global Engagement, IITM

  • Working with FORD Motors analytics on “Accelerated product design with domain knowledge augmented machine learning models”.

  • Currently developing a novel framework that combines Reinforcement learning and genetic algorithm for 3D design.

  • Developed a highly efficient 3D CNN algorithm to discover the errors in automobile parts. 

  • The developed algorithm is being tested and implemented in Volkswagen and FORD cars in the European Union.  

vcon 1.PNG

Research Associate , Vconnex Services

June 2020- March 2021

  • Worked in a Florida Based startup - Directly reporting to CEO

  • Developed a patient matching algorithm that overcomes the traditional patient record matching algorithm issues.

hpair%202_edited.jpg

Delegate, Harvard Project for Asian and International Relations

January 2021- January 2021

  • Selected as “one of the 300 Delegates ” from all over the world to attend the Harvard Project for Asian and International Relations (HPAIR), Harvard Conference 2021, which will take place from 15th-18th January 2021.

  • Part of HPAIR Alumni: Partnership with Asia Leader Series

Drone (1).png

Research Intern, Ericsson Research

January 2019- June 2020

Principal Investigator: Dr. Satheesh Kumar P and Dr. Saravanan Mohan

1.  Autonomous Drone for Mobile Tower Repairing and Monitoring. 

·         Designed and developed the world's first Autonomous Drone with a Robotic Arm on the top. 

·         Work published in International Conference of Control and Robotics. 

·         Presented in Ericsson Sweden. 

·         The developed drone is highly energy-efficient and saves the battery life up to 7 minutes. 

·         The drone is embedded with advanced computer vision and machine learning algorithms to identify the fault and carry out the                           repairing task in the Cell Tower. 

·         The drone is capable of autonomously moving in both Indoor and Outdoor Environment.

·         The repairing task is carried out with a robotic arm attached to the Drone which is in turn controlled by a novel stability algorithm. 

2.  A Novel Indoor Visible Light Communication to enhance data rate and reduce blockage.

·         Developed a robust mobility aware architecture for Visible light communication (VLC) in an indoor environment.

·         Work Published in ICCLCP, Sweden 2020

·         Work filed as a patent – US patent

·         A novel VLC localization technique was proposed to channelize the light according to user mobility

·         Developed a Multi-agent Reinforcement Learning algorithm to overcome the fundamental issue of blocking caused by people's                        shadow.

3. An Efficient model compression technique for developing energy-efficient deep learning architectures to reduce carbon footprint.

·         Developed a Context-aware Knowledge distillation technique with a novel loss function that enables light weight models to perform               efficiently on edge devices.

·         This technique won the 2nd prize in the International Telecommunication Union AI for Social good summit.

·         Work accepted as a Conference submission to ITU AI/ML in 5G challenge- AI for social good summit

·         Work accepted at IAICT

PP 1.jpg

Computer Vision Research Intern, Path Partners Technology

August 2019 - June 2020

Principal Investigator: Dr. Jayachandra Dakala, Mr. Praveen Nayak

1. Affective computing-based human emotion detection. 

  • Developed driver monitoring solution by applying advanced deep learning methodologies to assess drivers’ alertness under diverse conditions using a novel facial emotion recognition system decomposing Action units and valence and arousal.

  • The developed prototype was highly appreciated by the Indian Institute of Management, Bangalore and they partnered with us to develop the prototype into an actual product. 

  • Developed a system that used a hybrid combination of advanced facial analysis algorithms across 8 facial landmarks to perform real-time analysis of driver’s eyes, mouth, and head movements. (iBUG Computer vision Challenge).

  • Investigated the problem complexity and understanding the deep learning models’ abilities to develop algorithms that can run real-time on low power embedded systems.

  • Created a premise for emotional intelligence through a vision where the machine could not only interact but also empathize with people.

ROBOCON5.jpg

R&D Team Lead, Euroexim Bank Research

January 2019- December 2019

Autonomous Robots to work in coordination with Humans.

  • Lead a team of 16 undergraduates where we developed a Multiaxle (6-DOF) robotic ARM and an autonomous mobile robot that can assist in basic tasks and work in coordination with humans in proximity.

TCR2.jpg

Electrical Engineer , Technocrats Robotics- VITs Advanced Robotics team

July 12, 2023

  • Lead the embedded system subdivision of the team which comprised of 20 engineers to develop an intelligent robot for the Indian rover challenge 2019.

  • Fifth Place, IRC – Indian Rover Challenge- International Robotics competition- World Finals.

  • Third Place Nationally and got the special Mention since we were the first team to come in the top within 6 months of forming the team. 

  • Developed a next-generation MARS Rover with advanced communication and navigation systems which were designed to assist astronauts in MARS-like environments.

  • Designed the entire electronics subsystem and the differential drive system for the MARS Rover.

Ravel 1.jpg

Summer Intern, Ravel Electronics

April. 2018- May. 2018

Developed and deployed a prototype for peristaltic pumps and controlling the same with IoT using MQTT protocol and PLC to work with ease in nuclear reactors.

 
IRC new.jpg

Academic Research Experience

Learn to fail and innovate

logo_ucbridge_edited.png

Research Collaborator

Sep- 2021 to Present

- Quantifying and mining the rich information present in cellular features using GCNN to accelerate drug discovery

- Working on various representational learning algorithms to identify morphological patterns in cell populations resulting from chemical and genetic perturbations.

Learn More
Pathcheck.jpg

ML Researcher, Pathcheck Foundation, MIT Media lab

July 2020- Present

  • Working under the supervision of Prof. Ramesh Raskar to build robust, scalable and secure computational health
    models which secured global recognitions. (team work featured on Facebook, TechCrunch, wired etc)[Link]

  • Working with Camera Culture Group, MIT Media lab on Deep-ABM Models and VQ-VAE based image generation.

  • Part of the Research core Group (Published (including pre-prints) 14 papers, 3 more in pipeline)

  • Co-lead the team that participated in the Facebook data for challenge (Among 5 finalists globally)
    [Github][Github]

  • Working on several funding proposals with Prof.Ramesh Raskar in the Core Admin team of PathCheck Foundation.

  • Built deep learning based solutions for pandemic and was among the top 5 teams in FaceBook Data for Good and in the top 11 for the first round of XPRIZE Pandemic Response Challenge

  • Developed and deployed a rapid fact-checking search engine that curates the available information from various sources
    intelligently and summarizes the same to combat misinformation. (Solution for reliable COVID-19 news) [Search Engine

alan turing logo.jpg

Data Science Researcher

August 2021- October 2021

- Assisted in developing 4 models for predicting the functional relationship between epigenetic markers and DNA sequence.

- Investigated the receptive field of the Deepmind’s Enformer network. (Open Source Contribution)

RAN.png

Research Assistant, National Science Foundation-CNS core medium- WINLAB

September 2019- Present

  • Principal Investigator: Dr. Marco Gruteser, Dr. Kristin Dana, Dr. Narayan Mandayam

  • Developing a " Reality aware Networks" architecture that seeks to blend networking and sensing to develop a robust wireless architecture.

  • Exploring the design space of network architectures for blending perception and communications by designing low-energy tags and visual signaling strategies. 

  • Accelarting the progress by developing simultaneous localization and mapping algorithms that blend conventional strategies with network information to enhance robustness.

GSU 1.PNG

Research Asssitant, Georgia State University, CS Dept.- MORSE Studio

March 2019- Present

  • Principal Investigator: Prof. Ashwin Ashok

  • Worked on graphology based handwriting analysis system for normal and dyslexic people using Computer Vision and NLP. [Website]

  • Developed a novel form of visible light communication which increased the data rate by 24 times. [US patent pending]

  • Currently leading the Machine learning based camera communication research project. [Link]

  • Assisted on writing an NSF proposal on Machine learning assisted communication techniques.

drdo2.PNG

CAIR(Centre of Artificial Intelligence and Robotics), DRDO

Jan 2019- July 2019

  • Performed preliminary research and analysis with Dr. Reema Mathew on Energy efficient routing for wireless sensor networks.

  • Delivered my findings on efficient link quality prediction using Machine learning and engineered novel metrics.

ernet 2.PNG

ERNET- IIT Madras Research park

December 2019-June 2020

Worked with Dr. Pavendhan Arumugam- Director R&D ERNET, Ministry of Electronics and IT  and developed a novel Li-Fi based Indoor communication for a mobile user environment.

sankalp 2.png.jpg

UG Student Researcher

June 2016- June 2020


Developed efficient pulse compression and thinning techniques for a radar signal to dynamically coexist with well-established broadband spectrum DVB-T2(Sending to AEU Journal of Electronics and Communication Engineering).

 Designed a fault detection system with the dynamic reconfiguration of ECU in automobiles using FPGA.


Developed Deep Learning-Based compressed End-to-End Wireless Communication Systems With Conditional GANs as Unknown Channels for Video and audio data. 

 
aai%202_edited.jpg

Recognitions

Recipient of Fully funded NSF Master’s thesis Scholarship- Georgia State University- $55,190

February 2020

Recipient of Google Research Grant– $5000 for a research Proposed on AI for social good.

June 2020

Fifth Place, IRC – Indian Rover Challenge- International Robotics competition- World Finals

2020

Best Outgoing Student of the Department, VIT University 2016- 2020 – University Special Achiever

January 2018

First Place, VIT Robotics Makeathon- Robotics Competition-Regional Final.

2019

First Place, IEEE Robotics Makeathon-Robotics Competition-Regional Final.

June 2016

Special Mention, British Parliamentary Debate National Final

January 2018

First Place, CPCL Essay writing competition-National Final.

October 2013

Top 5 finalists among 132 teams globally in Facebook CMU UMD COVID challenge – Won $5000.

December 2020

Best Paper presentation award - ICCLCP Conference, Sweden

January 2018

Recipient of VIT University All India Entrance Examination Rank holders Scholarship- 100000Rs

Feature Subtitle

First Place, AAI Airport Authority of India-Elocution Competition-Regional Final.

June 2016

Best Speaker, International quality assurance council ( IQAC) ; On Quality of higher education.

January 2018

Special Mention, OWASP Ideathon–Pitching Regional Final

October 2013

Second Place, VIT Business Tactics Event-Regional Final.

June 2016

Third Place, KV Skit Competition-Regional Final

January 2018

Among Top 1% the country- 99% In Biology, All India Higher Secondary CBSE board examination.

June 2016

 

Patents

Georgia State University  - US Pat

Sethuraman T V, Susan Elias, Ashwin Ashok. “Machine learning Based M- ary Amplitude modulation for Visible Light Communication.”  (Filing in Progress)

Ericsson Research  - US Patent

Sethuraman T V*, Srinidhi Murali*, Pooja Sanathkumar*,Saravanan Mohan “Horizontal Switching method for Homogenous networking for Visible Light Communication using Multi-agent Reinforcement Learning ”,  PCT/IN2020/050748

Student Patent

Sethuraman T V et al, System and Method for assisting Graduate application assistance. (in progress)

 
ieee.png

Research Publication ( under construction )

Papers

 

Sethuraman T V, Vidya G, Thiruvikram G L, Satheesh K Perepu, Saravanan Mohan.” Energy Efficient knowledge distillation with a novel loss function”. Selected ITU Journal on Future and Evolving Technologies, 2020.

Abdulhaseeb Ahmed, Sethuraman T V et al. “An Empirical Study of Deep Learning Models for LED Signal Demodulation in Optical Camera Communication”, MDPI Network Journal- 2021.

Sethuraman T V, Kartik Singh Rathore, Amritha G, Kanimozhi G. “IoT based system for Heart Rate Monitoring and
Heart Attack Detection”. International Journal of Engineering and Advanced Technology (IJEAT) ISSN: 2249-8958, Volume-8 Issue-5, June 2019.

Vidya G, Thiruvikram G L, Sethuraman T V, Satheesh K Perepu, Saravanan Mohan.” Efficient knowledge distillation of teacher model to multiple student models”. IAICT 2021

Vishwath Kumar B S, Yogeeshwar S, Sethuraman T V, Velmathi G. “Deep learning based channel agnostic communication system for efficient data transmission”. Accepted in IEEE International Conference on Communication, Networks and Satellite 2020.

Pooja Sanathkumar*, Srinidhi Murali*, Sethuraman T V*, Saravanan Mohan, Pavendhan Arumugam, Ashwin Ashok. “Indoor VLC Channel Characterization for User Mobility: A Use-Case Study”. International Conference on Visible Light Communication and Positioning ICCLCP 2020. (Oral, Best Presentation award)

Dinesh Samuel Sathia Raj, Sethuraman T V, Yashvanth V P, Siva Shankar S, Satheesh K Perepu, Saravanan Mohan, “An Autonomous Drone with Robotic End Effector for Maintenance Operation of 5G Mobile Towers”, International Conference on Control Systems and Robotics-ICCR 2020 ( Best Presentation Award)

Sethuraman T V, Susan Elias, Ashwin Ashok “Demo: A machine learning based M-ary Amplitude Modulated visible light communication System”- 13th International Conference on COMmunication Systems & NETworkS COMSNETS 2020

Rohan Sukumaran*, Parth Patwa*, Sethuraman T V *, Sheshank Shankar, JosephBae, Rishank Kanaparati, Yash
Mathur, Abhishek Singh , Ayush Chopra , Sunny Kang , Ramesh Raskar, “COVID-19 Outbreak Prediction and
Analysis”. University of Michigan , AI symposium 2020 (Oral)

Darshan Gandhi, Sanskruti Landage,Joseph Bae, Sheshank Shankar,Rohan Sukumaran, Parth Patwa, Sethuraman T
V
, Priyanshi Katiyar, Shailesh Advani, Sunaina Anand, Aryan Mahindra, Rachel Barbar, Abhishek Singh, Ramesh
Raskar. Clinical Landscape of COVID-19 Testing: Difficult Choices. arXiv 2020

Rohan Sukumaran*, Parth Patwa*, Sethuraman T V*, Sheshank Shankar, Rishank Kanaparti, Joseph Bae, Yash
Mathur, Abhishek Singh, Ayush Chopra, Myungsun Kang, Priya Ramaswamy, Ramesh Raskar. COVID-19 Outbreak
Prediction and Analysis using Self-Reported Symptoms. Under Review AAAI Workshop, 2020.

Digital Landscape of COVID-19 Testing: Challenges and Opportunities - arXiv 2020

Challenges of Equitable Vaccine Distribution in the COVID-19 Pandemic – White Paper MIT Vaccine Conference

Projects

Worked on these to see how they work and fail :D

dr2.jpg
IRC.jpg

Autonomous drone for mobile tower repair

ROBOCON3.jpg
IITM%203_edited.jpg
LiFi robo.PNG
gan1.PNG
ITU.PNG
ctsan.PNG
vani app photo.PNG
HWGSU.PNG
ecg.PNG
Ml Arrhy 2.PNG
tata 1mg.PNG
ieee access paper pic.PNG
mdpi paper pic.PNG
IITM_Geosim_CV.PNG
pathpartners_pic.PNG
IMG-20200108-WA0011.jpeg
jdbs.PNG

Using self reported symptoms to estimate COVID-19 to compensate for scarcity in testing

Contact

Why should we communicate in 1/0s only?? Can we break the fundamentals?

Emotion Recognition using Deep learning( Valence Arousal, Action units, discrete emotions ).

Comparing similarity between two 3D objects

To what extent can deep learning aid in free air and underwater communication?

Can we detect similar looking objects where traditional vision based techniques fail?

Can online search trends help us be prepared for the next pandemic?

Multi Class Classification for cardiac arrhythmia.

IoT based system for Heart Rate Monitoring and Heart Attack Detection

Automated Graphology system which can estimate the users Psychological traits and mental health conditions( Dyslexia, Parkinson’s disease)

Android App for Early Detection of Dyslexia in children and providing customized learning pathways

GAN Based CT Scan Enhancement and Tumor detection by Generating a secondary CT from the primary counterpart.

Multistudent Knowledge Distillation- Compression of deep learning model

Deep learning based channel agnostic communication system for efficient data transmission

A robot that switches between LiFi and WiFi  and maintains  robust connectivity

Accelerated product design with domain knowledge augmented machine learning models.

Robots to work in the proximity of the humans for basic tasks - Human in the loop system

Details

MARS rover for assisting astronauts

 
sethu volunteer 1.jpg

Volunteer Work

Giving Back

aiforgood 2.PNG

Epidemic Modelling Research

September 2020 - Present

  • Developing a Simulator which is a web tool for NGOs and local authorities to model COVID-19 outbreak inside refugee camps and prepare timely and proportionate response measures needed to flatten the curve and reduce the number of fatalities.

  • This tool helps to predict the possible outbreak scenarios and their potential outcomes and help first responders design an optimal intervention strategy.

IMG-20201207-WA0008.jpg

VITeach - Social Outreach Club, Team Lead

April 2017 - Present

Lead the team and Conducted various social outreach activities for children across Tamilnadu and supported them technically with mid-day meal transport algorithms and assisted children with learning disabilities (dyslexia and autism).



Working on “participatory real-time digital solutions” to help kids with learning disability using voice analysis( Funded by Amazon India). 



Helped more than 483400 children by providing them a wide variety of resources to overcome financial, physical, and mental barriers and also helping them in getting access to the much-needed school education.



Developed an Optimized routing algorithm for the vehicles carrying  mid-day meals for poor children in Maharashtra and Gujarat which aided 300000+ children to get the mid-day meals on time. This is first where I felt I could utilize my technical skills as an engineer to contribute to the betterment of society.

mit vaccine.PNG

MIT Vaccine conference - Operations committee

Nov 2020- Present

  • Assisted Prof. Ramesh Raskar and MIT Media lab team to organize the MIT Vaccine conference. 

White File Folders

IEEE 5G World Forum, Publicity Committee Volunteer

May 2020–Sep 2020

Worked with Dr. Ashutosh Datta and team to enhance the outreach of the IEEE 5G world forum Conference. 

IRC 222.PNG

Mars Society south Asia-Indian rover Challenge - Assistant Head of Sponsorships

Aug 2019–Jan 2020

Lead the sponsorship team and procured 1.2 lakhs worth monetary funds and 50000 Rs worth non-monetary funds.

Vibrance.jpg

VIT Cultural and Sports Fest -Vibrance, Head of Public Relations and Point of Contact

Nov 2018–March.2019

Developed, implemented and evaluated the events marketing and communications strategy for advertisement, promotions, public relations and collateral to support the marketing objectives.

owasp 2.PNG

OWASP (Hacking & Security research group), Managment Team lead

JANUARY 2018 - MAY 2020

Gained experience in hacking from pen-testing on systems and organized events to spread awareness on cyber security.

Get in Touch
hackhub.png

IEEE Computer Society, Research and development team

JUNE 1028 - MAY 2020

  • Revamped the society's focus on software engineering and building network on and off campus.

  • Proposed various marketing strategies and networking activities to bring in generic problem statements and funds to solve the same.

VIThink new.PNG

VIThink (ML and AI Research Group) , AI Researcher

April 2017 - May 2020

Conducted workshop, events to spread AI awareness and carried out research activities on supporting orphan and poor kids.

technovit.jpg

VITs National Technical Fest - TechnoVIT-Marketing Core Team

July 12, 2023

This is a brief and concise description of your volunteer experience. Use this space to describe the philanthropic activity, the population it supported, as well as your specific role, impact and contribution.

rennaisance 1.jpg

Renaissance ‘19 (National Entrepreneurship fest) , Core committee

June 2018 - Feb 2019

Organized the fest being a part of Operations, Marketing and by heading Logistics and Partnership committee.

 
Robowar.jpg

My Skills

Tricks of the Trade

 

Cloud Computing

Worked with AWS and Google Cloud Platform. Build and maintained applications that are research and product perspective.

Docker

As a software engineer, I have contributed in making multiple dockersied applications. Have also created out of the box embedded applications using docker.

Linux

Been using as the main operating system for more than a year now. Capable of writing bash/shell scripts to automate daily tasks.

Git

Experienced with Git-based development and protocols. Worked with github and gitlab.

C/C++

Experience with C and C++ programming. Used it for writing programming contest, problem solving and light weight models.

Python

Well experienced with python and building applications for over 4 years. Have used it to develop machine learning models, optimized code blocks and flask websites.

lua

Able to make weight API servers using lapis framework for small edge devices. Have utilized this in Drone navigation codes.

MATLAB

Worked on Digital Signal Processing, Image processing and control systems modules. Well experienced with SIMULINK

Data Engineering and analysis

This involves few of my work in Ericsson Research and Pathcheck, MIT. I have also worked with Databricks notebooks, flintrock and other data tools like pyspark, HDFS along with AWS S3.