Thabsheer Jafer Machingal

Machine Learning Researcher @ Apex Microdevices. MS in Robotics from WPI.

I am a Machine Learning Researcher at Apex Microdevices, specializing in inverse design and topological optimization for metasurfaces and metalenses.

I have obtained a master's in Robotics Engineering from Worcester Polytechnic Institute in 2023.

My research interests include Robot Learning, Perception, and Artificial Intelligence, with an emphasis on integrating intelligent systems into robotics.

To know more about me, you can check out my CV.

Research

Research Engineer - Inverse Design and Topological optimization using Machine Learning
Advisor : Dr.Debanik Das
Apex Microdevices

My work focuses on employing advanced deep learning techniques—including neural networks, encoder-decoder models, and generative adversarial networks (GANs)— to explore exhaustive parametric spaces for the design of metasurfaces. By building surrogate models that approximate Maxwell's equations of electromagnetism, I aim to accelerate the design process in photonics

Master's Research: Object Rearrangement for Robotic Waste Sorting
Manipulation & Environmental Robotics Lab
Worcester Polytechnic Institute
pdf | poster

Experience

Apex Microdevices Feb 2024 - current

Machine Learning Research Engineer

Agot Co. May 2023 - Aug 2023

Computer Vision Intern

Education

Worcester Polytechnic Institute Jan 2022 - Dec 2023

Research Scientist - Machine Learning

Kerala Technological University Aug 2017 - May 2021

Mechanical Engineering

Relevant Projects

Dynamic Obstacle Avoidance & Path Planning in a Hospital Environment
|

The "Dynamic Obstacle Avoidance and Path Planning in a Hospital Environment" project evaluates three algorithms: D*, Splines with Dynamic Window Approach (DWA), and Hybrid Potential-Based Probabilistic Roadmap (HPPRM). These methods are tested in dynamic hospital settings, focusing on safe and efficient navigation. D* adapts to changing obstacles, HPPRM generates smooth paths, and Splines+DWA optimizes real-time control. The project identifies the most suitable approach for hospital robotics based on safety and computational efficiency.

Dominic Ferro, Thabsheer Jafer Machingal, Girivaasan Chandrasekaran
Worcester Polytechnic Institute

Semantic Mapping: Pointcloud painting

The "Semantic Mapping for Autonomous Robots" project focuses on enabling robots to map environments with semantic understanding. Using deep learning and SLAM (Simultaneous Localization and Mapping), the robot can recognize and label objects while creating a spatial map. This approach enhances autonomous navigation by providing context-aware mapping, which improves decision-making in complex environments.

Thabsheer Jafer Machingal
Worcester Polytechnic Institute

NeRF: Neural Radiance Field

The "NeRF: Neural Radiance Fields" project implements a method for generating 3D scenes from 2D images using neural networks. By optimizing the volumetric scene representation, NeRF produces highly realistic views from different angles. This project enhances image-based rendering, enabling the creation of 3D models with fine details and complex lighting.

Thabsheer Jafer Machingal, Krishna Madhurkar
Worcester Polytechnic Institute

My Autopano: Image Stitching

The "My AutoPano" project automates panoramic image stitching by detecting and aligning features from multiple photos. Using computer vision techniques, it seamlessly merges overlapping images into a wide-view panorama. The project simplifies the process of generating panoramic images while ensuring smooth transitions between individual photos.