IEEE/RSJ IROS WS — Geometric Representations
in Robotics
Structured priors, metrics, constraints &
groundings for Manipulation
Call for Papers
We invite extended abstracts and position papers for interactive poster presentations at the IEEE/RSJ IROS 2026 Workshop on Geometric-Aware Representations in Robotics.
Important Dates
- Call for Participation Opens: TBD 2026
- Submission Deadline: TBD 2026
- Notification of Acceptance: TBD 2026
- Camera-ready Submission: TBD 2026
- Workshop Date: TBD, — IROS 2026
Awards
- Best Student Paper Award – Recognizing outstanding student contributions to the field of geometric methods in robotics
- Best Video / Demo Award – Recognizing impactful demonstrations of geometric representations in robotic manipulation
Submission Instructions
Please submit an extended abstract or position paper of 2 to 4 pages in IEEE format (excluding references) via the workshop's submission system (link coming soon). Submissions may also include supplementary video material, demos, or visual content, especially when illustrating geometric structure, contact-rich interaction, or representation-driven manipulation behaviour.
All submissions will undergo peer review by at least two domain experts, with attention to:
- Relevance to the workshop themes
- Technical quality and novelty
- Potential to stimulate discussion during the workshop
Accepted contributions will be presented through spotlight teaser talks and interactive poster sessions, giving contributors visibility and direct access to feedback from invited speakers and attendees. Outstanding accepted authors may also be invited to participate in the Geometric Representation Interdisciplinary Impact Clinics (GRIIC) for live mentoring from senior experts.
Topics of Interest
We welcome contributions on (but not limited to) the following topics:
- Geometric priors and groundings for sensing and state estimation on manifolds (SE(3), SO(3), etc.)
- Equivariant and symmetry-aware models for robot perception and manipulation
- Structure-preserving learning and manifold-valued motor skills
- Learning on Riemannian manifolds and Lie groups
- Geometry-aware motion planning and trajectory optimization
- Geometric mechanics and contact-rich manipulation
- Task-space control with geometric metrics and constraints
- Geometric world models, affordances, and generative representations
- Object-centric and scene-centric representations for manipulation
- Geometric foundations for whole-body and dexterous manipulation
- Composability and generalization through geometric structure
- Applications: dexterous manipulation, bimanual tasks, human-robot interaction
We especially encourage submissions from women, early-career researchers, and interdisciplinary teams. Submissions that connect traditionally separate communities (e.g., geometric control + robot learning, tactile sensing + manifold methods) are particularly welcome.