About Me
Hello! I am a Research Scientist at Georgia Institute of Technology working on Activity Recognition using Machine Learning. I have enjoyed working on a range of projects in Natural Language Processing, Computer Vision and Robotics in both industrial and academic setups.
I graduated with a Masters in Computer Science from GeorgiaTech where I specialized in Machine Learning, NLP and CV. I completed my Bachelor’s degree in Computer Science from BITS Pilani where I took up courses and projects related to Pattern Recognition, Machine Learning and Computer Graphics. Through my interships and work, I have had the opportunity to implement projects in these fields.
More details about my publications, projects and experience below.
Publications
PBP: Path-based Trajectory Prediction for Autonomous Driving [arxiv]
Sepideh Afshar, Akshay Bhagat, Nachiket Deo, Titas Chakraborty, Balarama Raju Buddharaju, Adwait Deshpande, Henggang Cui
Accepted to ICRA 2022
Indic-Transformers: An Analysis of Transformer Language Models for Indian Languages [arxiv]
Kushal Jain *, Adwait Deshpande *, Kumar Shridhar, Felix Laumann, and Ayushman Dash
Accepted to ML-RSA workshop, NeurIPS 2020
NERDSearch : Using Familiar Gameplay Mechanics for Linguistic Annotation [pdf]
Adwait Deshpande and Nischay Ghattamraju
Accepted to Wordplay, NeurIPS 2020
Experience
GeorgiaTech, Atlanta, GA
Research Scientist II, AI-CARING (Oct 2023 - Present)
Developing state-of-the-art activity recognition for smart homes to assist the elderly in daily living
Link
Motional, Pittsburgh, PA
Machine Learning Intern, Prediction (May 2022 - Aug 2022)
R&D for lane-aware and map complian motion prediction models based on Transformers
Reverie Language Technologies Pvt. Ltd., Bengaluru
Technical Lead, R&D (July 2019 - July 2021)
NLP for low corpus Indian Languages to solve problems in digital accesibility
Endless Robotics Pvt. Ltd., Hyderabad
Computer Vision Engineer / Software Architect (Aug 2016 – Apr 2019)
Responsible for desigining the software architecture for a one-of-a-kind wall painting robot
Researched and developed 3D mapping, image rectification, pattern recognition to help the robot navigate in sparsely featured environments
Tonbo Imaging Pvt. Ltd., Bengaluru
Member of Technical Staff, R&D (July 2015 – July 2016)
Real-time Multi-object detection and tracking for visible and IR video streams
Tonbo Imaging Pvt. Ltd. Bengaluru
Intern (July 2014 - Dec 2014)
Developed application to improve accuracy and cut down time required for camera calibration
Worked on OpenSceneGraph for rendering configurable, realistic video scenes for data generation
Central Electronics Engineering Research Institute, Pilani
Project Assistant, Study Oriented Project (Jan 2014 - Apr2014)
Worked on head detection in different poses using cascaded classifiers
Projects
Unsupervised clustering for sign languages more
3D map for sparsely featured environments
Project at Endless Robotics
code
Multi object detection and tracking
Project at Tonbo Imaging Pvt. Ltd.
video
Head detection using cascaded classifiers
Project at CEERI
code
Image rectification using AprilTags
Project at Endless Robotics
code