Visuotactile Representation Learning for Robotic Manipulation - Honda Research Institute USA

Visuotactile Representation Learning for Robotic Manipulation

Your application is being processed

Visuotactile Representation Learning for Robotic Manipulation

Job Number: P24INT-07
We are seeking passionate individuals who will contribute to the development of cutting-edge perception algorithms, integrating computer vision and tactile sensing (visuotactile) to enhance the capabilities of robots in in-hand object manipulation. This project focuses on developing innovative perception algorithms to estimate object states and finger-object contact states using visuotactile information and learning tactile representations that generalize across tasks. The successful candidate will have the opportunity to contribute to the development of these algorithms in simulation and deployment on advanced robotic platforms with multi-fingered hands and ensure their performance on the robot matches that of the simulations.
San Jose, CA

 

Key Responsibilities

 
  • Develop algorithms to fuse vision and tactile data for perceiving finger-object contact conditions.
  • Develop generalizable tactile representations applicable across various tasks.
  • Explore temporal approaches to improve the stability of the estimates over time.
  • Implement and validate the developed algorithms in simulation and on hardware.

 

Minimum Qualifications

 
  • Ph.D.  or highly qualified M.S. candidate in computer science or a related field.
  • Experience in deep learning and other machine learning methods.
  • Good programming skills in Python.
  • Experience with Robot Operating System (ROS).

 

Bonus Qualifications

  • Experience with sim-to-real approaches.
  • Experience with deep learning approaches such as GNNs and LSTMs.
  • Experience in perceiving object properties using force and pressure data from tactile sensors.
  • Experience with optical tactile sensors like DIGIT
  • Experience with simulation tools such as IsaacSim
  • Experience with PyTorch

 

Years of Work Experience Required  0
Desired Start Date 5/5/2025
Internship Duration 3 Months
Position Keywords Vision and Tactile Contact Detection, Robot Operating System, Machine Learning

 

  

Alternate Way to Apply

Send an e-mail to careers@honda-ri.com with the following:
- Subject line including the job number(s) you are applying for 
- Recent CV 
- A cover letter highlighting relevant background (Optional)

Please, do not contact our office to inquiry about your application status.