MIT EECS Angle Undergraduate Research and Innovation Scholar
Electrical Engineering and Computer Science
Object Detection for Robotic Systems Our goal for this project is to develop an object detection system for robotic systems using computer vision techniques. Robotic systems use object recognition systems to build maps of their environment which are subsequently used in various aspects of autonomous control such as path planning and localization. We plan to apply a full scene segmentation approach to do object detection. As applied to object detection image segmentation allows the robot to recognize objects of interest in the input visual feed and pinpoint the exact pixel locations of these objects. An autonomous robot navigating indoors would for example be able to perform segmentation on its camera feed to determine pixels of the image corresponding to the ground walls and other obstacles such as tables and chairs.
I became interested in robotics and computer vision while taking 6.141 and doing research with Prof. Torralba's group last spring and have worked on machine learning and deep learning projects at previous internships. I hope to get some meaningful research experience through the SuperUROP program.