Research Engineer
Avataar
Georgia Institute of Technology
Email / LinkedIn

Do we settle for the world as it is, or do we work for the world as it should be?

— Barack Obama

News

  • [Feb 25] Promoted to RE III at Avataar
  • [Jun 24] Attended CVPR in Seattle.
  • [Feb 24] Our work on Domain Adaptation using videos got accepted at TMLR.
  • [Oct 23] M3L accepted at WACV 2024.
  • [Jul 23] Started as a Research Engineer II at Avataar.ai
  • [Jun 23] Moved back to India! Read more here.
  • [May 23] Graduated with an MS in CS from Georgia Tech.
  • [May 22] Started summer internship at Amazon Fashion as an Applied Scientist.
  • [Jan 22] Working with Prof Zsolt Kira on label-efficient frameworks.
  • [Nov 21] ‘Recommendation of Compatible Outfits Conditioned on Style’ accepted for presentation for the ECIR’22 Full Paper track!
  • [Oct 21] Fashion Compatibility accepted for demo at CODS-COMAD 2022!
  • [Aug 21] Working as Graduate Research Assistant with Prof. Devi Parikh.
  • [Aug 21] Starting MS in CS (Specialization in Machine Learning) at Georgia Tech.
  • [Jul 19] Starting as a Data Scientist at Flipkart responsible for recommendations.

Bio

I am a Research Engineer at Avataar.ai where I work on generative AI and 3D with Dr. Sohil Shah and Dr. Shubham Goel. I did my MS from the department of Computer Science at Georgia Institute of Technology, advised by Prof. Zsolt Kira and Prof. Devi Parikh. Before that I was a Data Scientist at Flipkart in the Recommendations team supervised by Dr. Aditya Rachakonda, Dr. Arnab Bhattacharya and previously by Samik Datta. Prior to joining Flipkart, I completed my undergrad at IIT Delhi advised by Dr. Prathosh AP.

Broadly my interest lies in Computer Vision, Vision + Text, Creative AI. I strongly believe that technology benefits human society and has a lot of potential to improve human lives.

For about an year, I also volunteered my extra time with DSIndiaVsCovid mentored by Dr. Srujana Merugu, Dr. Mohit Kumar and Dr. Alpan Raval in collaboration with Wadhwani AI to help government agencies in India in their covid response.

My CV is available here.

Research - Avataar
2023 - Present
MS - Georgia Tech - Computer Science
2021 - 2023
Applied Scientist Intern - Amazon Fashion
Summer 2021
Data Scientist Volunteer - Covid 19 Data Science Consortium with Wadhwani AI
2020 - 2021
Data Scientist - Flipkart - Recommendations
2019-2021
B.Tech - IIT Delhi - Electrical Dept.
2015 - 2019

Publications and Preprints

* Equal contribution

Missing Modality Robustness in Multi-Modal Semi-Supervised Semantic Segmentation

Harsh Maheshwari, Yen-Cheng Liu, Zsolt Kira
WACV'24
Project Page arXiv Code

Performance degradation of ImageNet trained models by simple image transformations

Harsh Maheshwari
arXiv only
arXiv Code

Recommendation of Compatible Outfits Conditioned on Style

Harsh Maheshwari*, Debopriyo Banerjee*, Lucky Dhakad*, Muthusamy Chelliah, Niloy Ganguly, Arnab Bhattacharya
ECIR'22 (oral)
CODS-COMAD 2022 Demo track
Project Page ECIR Paper arXiv CODS-COMAD Demo Paper

CoSIR: Optimal control of SIR epidemic dynamics by mapping to Lotka-Volterra System

Harsh Maheshwari, Shreyas Shetty, Nayana Bannur, Srujana Merugu
ICLR'21 MLPCP Workshop, CHIL'21 Workshop
Paper Code Demo Poster

A Flexible Data- Driven Framework for COVID-19 Case Forecasting Deployed in a Developing- world Public Health Setting

Sansiddh Jain, Avtansh Tiwari, Nayana Bannur, Ayush Deva, Siddhant Shingi, Vishwa Shah, Mihir Kulkarni, Namrata Deka, Keshav Ramaswami, Vasudha Khare, Harsh Maheshwari, Soma Dhavala, Jithin Sreedharan, Jerome White, Srujana Merugu, Alpan Raval
AI for Social Impact book
Paper Complete Book

Adaptive COVID-19 Forecasting via Bayesian Optimization

Nayana Bannur, Harsh Maheshwari, Sansiddh Jain, Shreyas Shetty, Srujana Merugu, Alpan Raval
CoDS-COMAD 2021
Paper Code

Audience Creation for Consumables - Simple and Scalable Precision Merchandising for a Growing Marketplace

Shreyas S*, Harsh Maheshwari*, Avijit Saha*, Samik Datta*, Shashank Jain, Disha Makhija, Anuj Nagpal, Sneha Shukla, Suyash S
Paper


(Courtesy: Abhishek Das)