Bridging Vision, Language, and Action: What's Missing in Actionable Visual Perception for Robotics

Submission Instructions

We welcome submissions covering:

• Research papers: Long papers (8 pages) showcasing novel findings, methods, or theoretical advancements.
• Short/Abstract papers: Features exploratory work (4 pages or 2 pages excluding references) that may be preliminary but presents innovative concepts, early results, or thought-provoking viewpoints that stimulate discussion and future work.
• Position papers: Offer critical perspectives on trends and challenges within the filed (no less than 8 pages).
• Survey papers: Provide thorough reviews of specific topics, mapping the current research landscape and suggesting directions for future exploration (no less than 8 pages).

All formats allow unlimited references and appendices.

Contributions will be non-archival but hosted on our workshop website, and thus dual submission is allowed where permitted by third parties. We welcome submissions that are under submission or accepted by other conferences. Please mention it in the last sentence of the paper abstract if your paper has been under submission or accepted by other conferences.

Submissions should follow CVPR two-column style and be anonymous; see the CVPR-26 author kit for details.

Submission and Important Dates

• Portal: OpenReview Submission Portal
• Start Date: March 9th, 2026
• Deadline: May 14th, 2026 (23:59PM AOE)
• Notification to Authors: May 24th, 2026 (23:59PM AOE)
• Camera-ready Due (non-archival): May 28th, 2026 (23:59PM AOE)

To encourage open-ended discussion and maximize in-person engagement, the workshop will feature a mix of structured and interactive formats.

• Keynote sessions will consist of 40-minute presentations followed by 5-minute Q&A segments, providing speakers with the opportunity to engage directly with attendees.
• Poster sessions and Coffee breaks will showcase recent advancements and novel research while also serving as a platform for networking and idea exchange among researchers from diverse backgrounds and career stages.
• Panel discussion will incorporate live polling and audience-submitted questions, fostering dynamic and meaningful dialogue. We will invite leading researchers from both academia and industry to participate in a debate on the most promising directions for future research in embodied AI from a computer vision perspective.

These interactive elements are designed to stimulate lively exchanges, bridge the gap between junior and senior researchers, and cultivate an open, inclusive research community.

1. Multi-Objective Photoreal Simulation (MOPS) Dataset for Computer Vision in Robot Manipulation
2. SIR: Structured Image Representations for Explainable Robot Learning
3. Learning from Trials and Errors: Reflective Test-Time Planning for Embodied LLMs
4. SCALE: Self-uncertainty Conditioned Adaptive Looking and Execution for Vision-Language-Action Models
5. LightSplat: Fast and Memory-Efficient Open-Vocabulary 3D Scene Understanding in Five Seconds
6. Dex4D: Task-Agnostic Point Track Policy for Sim-to-Real Dexterous Manipulation
7. Scaling Verification Can Be More Effective than Scaling Policy Learning for Vision-Language-Action Alignment
8. PhaForce: Phase-Scheduled Visual–Force Policy Learning with Slow Planning and Fast Correction for Contact-Rich Manipulation
9. PhysMem: Scaling Test-Time Memory for Embodied Physical Reasoning
10. Multimodal Causal Subtask Modeling for Scalable VLA Pipelines in Long-Horizon Manipulation
11. Learning End-to-End Visuomotor Control via Stochastic Decoupled Policy Gradient
12. VL-Nav: A Neuro-Symbolic Approach for Reasoning-based Vision-Language Navigation
13. PointWorld: Scaling 3D World Models for In-The-Wild Robotic Manipulation
14. Towards Robust Robot Manipulation: Visual-Tactile Sensing for VLA Models
15. Geometry-Regularized Affordance Prediction for Sim-to-Real Transfer in Vision-Language-Action Models
16. STORM: Slot-based Task-aware Object-centric Representation for robotic Manipulation
17. Dream2Flow: Bridging Video Generation and Open-World Manipulation with 3D Object Flow
18. Artiverse: A Diverse and Physically Grounded Dataset for Articulated Objects
19. ULTRA: Unified Multimodal Control for Autonomous Humanoid Whole-Body Loco-Manipulation
20. KEYGEN: Unsupervised Keypoint based Object-Centric Representations for Category-Level Policy Generalization
21. HoMMI: Learning Whole-Body Mobile Manipulation from Human Demonstrations
22. Learning Humanoid End-Effector Control for Open-Vocabulary Visual Loco-Manipulation
23. Point and Check: Diagnosing Fragile Visual Grounding for Embodied Systems
24. VLS: Steering Pretrained Robot Policies via Vision–Language Models
25. APPLV: Adaptive Planner Parameter Learning from Vision-Language-Action Model
26. Moving Through Clutter: Scaling Data Collection and Benchmarking for 3D Scene-Aware Humanoid Locomotion via Virtual Reality
27. Multi-Phase Vision-Based Navigation and Inspection for Legged Robots with Online Goal Refinement and Vision-Only Halting
28. ForesightNav: Learning Scene Imagination for Efficient Exploration
29. Before the Needle: Action-Conditioned Risk-Affordance Maps for Safe Ultrasound-Guided Intervention
30. Can Vision-Language Models See the Future? Coverage-Aware Pedestrian Forecasting and Intent-Guided Residuals
31. From Vision to Action: Benchmarking VLMs for Multi-Arm Robotic Fruit Harvesting
32. Text to Multi-Component Robotic Assembly: Vision Language Reasoning for Functional Component Assignment
33. Reasoning-Guided Part-Level Visual Grounding via Reinforcement Learning
34. CANON: Canonical Observation Pretraining for Visuomotor Control