Kate Uzar

Research project title: 
Learning to grasp movable objects based on tactile information
Research project: 

Current Project Title: Real-time vision and tactile sensor fusion for object recognition.

Object recognition approach that uses tactile (touch) data for initial recognition and visual information only for the confirmation of the recognised object. This project builds upon the existing framework for visual and tactile sensor fusion but changes the order the sensory information is used. Most literature approaches focus on approaches where visual information is used first for recognition process, however, an inspiration based on how humans recognise objects suggest that the sense of touch is used from the very beginning of a recognition process when a new object is encountered.

The project is funded by Honda Research Institute.

About me: 

I am a first year PhD student at the Edinburgh Centre for Robotics under a supervision of Prof. Sethu Vijayakumar and Dr. Michael Gienger. I hold a BSc(Hons) in Computer Science with a placement year. I'm interested in applying Machine Learning techniques to robotics.