Hi! I'm a Machine Learning (AI) Engineer with 6+ years of full-time software development experience, in which over 4 years are in CV/ML Research and Development. I'm currently an AI Software Engineer at Musashi AI. Before that, I was a Post-graduate Research Associate at Bell Multimedia Laboratory where I worked in eXplainable Artificial Intelligence (XAI) in Computer Vision, along with LG AI Research. I received my Master's degree in Computer Vision and Robotics from the Department of Electrical and Computer Engineering (ECE) at the University of Toronto (Class of 2020).
My interests broadly lie in 2D and 3D Computer Vision, Efficient and Interpretable Deep Learning, Industrial Automation, Robotics and related applications. Currently I'm working on developing and deploying efficient AI solutions in the edge devices for Visual Quality Inspection across various automotive OEMs.
Experience
- AI Software Development Engineer, Musashi AI Jan 2021 - Present
- Led the development of a real-time instance segmentation model that achieved high detection accuracy at the micron level.
- Initiated and led the creation of scalable Inspection Software that can be easily replicated for a diverse clientele, thereby enhancing the company's product portfolio.
- Designed customer-centric dashboards that offer comprehensive insights into AI systems, resulting in a standalone product.
- Successfully deployed AI solutions across cloud platforms and edge devices in industrial settings.
- Established effective communication architectures between PLC, Robot, and other 2D/3D vision sensors for efficient data transfer.
- Worked extensively with FANUC Cobots and integrated solutions with leading PLC providers like Mitsubishi, Keyence, Allen Bradley, and others.
- Specialized in optical equipment, 3D profile sensors, and computational imaging for metallic automobile components.
Waterloo, ON
- Research Associate, LG AI Research (Academic partnership with UofT) Sep 2019 - Feb 2021
- Post-graduate Researcher at Bell Multimedia Laboratory (ECE, University of Toronto) under the supervision of Prof. Konstantinos (Kostas) N. Plataniotis in collaboration with LG AI Research (South Korea) on a year-long initiative to interpret deep residual ML models.
- Proposed, developed, and patented a novel eXplainable AI (XAI) algorithm, "Semantic Input Sampling for Explanation" (SISE), now integrated into LG’s industrial codebase for fully automated supervision of LG Display Co. Ltd. Read the official blog post by LG in Korean.
- Published and presented 3 Academic Research Papers at top-tier AI conferences - AAAI'21 and IEEE ICASSP'21.
Toronto, ON
- Systems Engineer, Infosys May 2016 - Jul 2018
- Backend developer for AIMIA (formerly Groupe Aeroplan), a data-driven loyalty analytics company based in Montreal.
- Developed and delivered multiple stored procedures in the MySQL relational DB management system server.
Bangalore, India
Publications
- Ada-SISE: Adaptive Semantic Input Sampling for Efficient Explanation of Convolutional Neural Networks Mahesh Sudhakar, Sam Sattarzadeh, Konstantinos N. Plataniotis, Jongseong Jang, Yeonjeong Jeong, Hyunwoo Kim
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021
[paper] [arxiv] [poster] [slides] [bibtex]
- Integrated Grad-CAM: Sensitivity-Aware Visual Explanation of Deep Convolutional Networks via Integrated Gradient-Based Scoring Sam Sattarzadeh, Mahesh Sudhakar, Konstantinos N. Plataniotis, Jongseong Jang, Yeonjeong Jeong, Hyunwoo Kim
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021
[code] [paper] [arxiv] [poster] [slides] [bibtex]
- Explaining Convolutional Neural Networks through Attribution-Based Input Sampling and Block-Wise Feature Aggregation Sam Sattarzadeh, Mahesh Sudhakar, Anthony Lem, Shervin Mehryar, K. N. Plataniotis, Jongseong Jang, Hyunwoo Kim, Yeonjeong Jeong, Sangmin Lee, Kyunghoon Bae
Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), 2021
[news] [paper] [arxiv] [poster] [slides] [bibtex]
- SVEA: A small-scale benchmark for validating the usability of post-hoc explainable AI solutions in image and signal recognition Sam Sattarzadeh, Mahesh Sudhakar, Konstantinos N. Plataniotis, Jongseong Jang, Yeonjeong Jeong, Hyunwoo Kim
Proceedings of the IEEE/CVF International Conference on Computer Visiong (ICCV), 2021
[paper] [bibtex]
IP & Patents
- Computer system and method for automated visual inspection using segmentation-based anomaly detection Martin Bufi, Mahesh Sudhakar, Ross Tsenov
Musashi AI North America Inc., 2023
[IP] [pdf]
- Computer system, method, and device for active learning Martin Bufi, Saeed Bakhshmand, Ross Tsenov, Mahesh Sudhakar, Jasmeen Kaur
Musashi AI North America Inc., 2023
[IP] [pdf]
Education
- Master of Engineering, Specialization in Robotics and AI University of Toronto, Electrical and Computer Engineering (ECE), Sep 2018 - Apr 2020
GPA: 3.97/4.0
- Bachelor of Engineering, Electrical and Electronics Engineering (EEE) Velammal Engineering College, Anna University, India, 2016
Cumulative GPA: 8.54/10.0