Research Interests
Embodied AI, Autonomous Navigation, Robotic Manipulation, Computer Vision, Machine Learning
As a student researcher, Mahtab focuses on building embodied robotic systems capable of autonomous navigation and interaction in real-world environments, with an emphasis on perception-driven decision making and control.
Research Focus
His primary research investigates autonomous indoor navigation and mobile manipulation, integrating computer vision, sensor fusion, and learning-based methods for localization, mapping, path planning, and obstacle avoidance. He works with imitation learning and reinforcement learning to develop end-to-end navigation and control policies and deploy them on physical robots with tight hardware-software integration.
In parallel, he explores robotic manipulation and applied computer vision, including visual-language-guided manipulation and Document AI pipelines (OCR, VLMs) for structured information extraction.
His broader goal is to translate state-of-the-art AI research into robust, deployable robotic systems that operate reliably in unstructured, real-world settings.