Towards Accessible and Intuitive Shared Autonomy
| Type of publication: | Conference paper |
| Citation: | Rajapakshe_HRI'25-2_2025 |
| Publication status: | Published |
| Booktitle: | Proceedings of the 2025 ACM/IEEE International Conference on Human-Robot Interaction |
| Year: | 2025 |
| Month: | March |
| URL: | https://dl.acm.org/doi/10.5555... |
| Abstract: | The deployment of assistive robotics technologies in human environments is often hindered by the diverse personalization requirements of individual users. A promising solution to this challenge is shared autonomy, enabling collaboration between robots and users. The goal of my PhD is to develop shared autonomy methods that empower users with disabilities to control assistive robots efficiently in their daily lives. Specifically, I aim to address the dimensionality gap encountered when using simpler interfaces while maintaining an intuitive input mapping. Our prior work introduced a geometric shared autonomy approach based on canal surfaces and a dynamic input mapping framework, that serve as foundational efforts in this area. To gather preliminary feedback, I conducted a user study involving wheelchair users, utilizing a specialized joystick designed for accessible video gaming. In future work, I plan to incorporate computer vision to eliminate reliance on demonstrations and use participatory design to enhance our interface design. |
| Main Research Program: | Human-AI Teaming |
| Additional Research Programs: |
AI for Everyone |
| Keywords: | |
| Projects: |
Idiap |
| Authors: | |
| Added by: | [UNK] |
| Total mark: | 0 |
|
Attachments
|
|
|
|
|
Notes
|
|
|
|
|
