log in  |  register  |  feedback?  |  help  |  web accessibility
Logo
PhD Proposal: Grasping in The Dark: Compliant and Adaptive Grasping with Tactile Feedback
Kanishka Ganguly
Remote
Wednesday, January 6, 2021, 1:00-3:00 pm Calendar
  • You are subscribed to this talk through .
  • You are watching this talk through .
  • You are subscribed to this talk. (unsubscribe, watch)
  • You are watching this talk. (unwatch, subscribe)
  • You are not subscribed to this talk. (watch, subscribe)
Abstract
Robust and adaptive grasping is still an unsolved problem in the field of robotics. In order to solve this problem, tactile feedback is a necessary requirement, especially when anthropomorphic grippers are concerned. In this proposal, we deal with some of the problems concerning grasping with high degree-of-freedom grippers such as the Shadow Dexterous Hand, and use the SynTouch BioTac sensors as tactile sensors. We present the concept of computational tactile flow, by combining various models of the sensor with traditional computer vision based optical flow techniques. We also discuss the process of integrating tactile feedback into a grasping pipeline using the ShadowHand, and provide a feedback control system for adaptive grasping. This is useful in solving the problem of grasping in situations where visual feedback is limited or absent.

We discuss ongoing work on trying to generate spatio-temporal events from the BioTac sensor data and how interpreting the existing data in both spatial and temporal dimensions allows for easier analysis of motion across the sensor surface and makes it more robust to noise. Lastly, we present future work on teleoperation of arm and Hand for data gathering, and eventual applications of this data to learning based techniques for end-to-end control of the system for grasping and surface exploration.

Examining Committee: 
 
                          Chair:               Prof. Yiannis Aloimonos      
                          Dept rep:         Prof. Dinesh Manocha
                          Members:         Dr. Cornelia Fermuller 
Bio

 

Kanishka Ganguly is a PhD student of Computer Science, advised by Prof. Yiannis Aloimonos. He has done his Masters of Engineering in Robotics from UMD in 2017. Kanishka works on dexterous grasping with the ShadowHand at the Perception and Robotics Group, and his research focuses on integrating tactile feedback with visual perception to develop a tactile-perceptual theory of grasping.
This talk is organized by Tom Hurst