mlsanityData to 14 Oct 2026

CrossTracer: Cross-Embodiment Navigation via VLA Model Reasoning and Trace Residuals Adapting

arXiv 2608.06688 ↗First posted 7 Aug 2026cs.ROcited by 0

On the map: Vision-language-action models → Human-video and latent-action pretraining

Abstract (arXiv)

Vision-language-action (VLA) models provide strong semantic priors for robot navigation, but they often ignore embodiment-specific mobility constraints. A path that is semantically plausible for one robot may be physically infeasible for another. We propose CrossTracer, a hierarchical framework for cross-embodiment navigation through adaptive trace residuals. CrossTracer represents navigation plans as normalized image-plane waypoints, forming a unified pixel-space interface between semantic reasoning and physical grounding. First, Vision-Language Trace Proposer (VL-Tracer) adapts a pretrained VLA model to predict an initial navigation trace from egocentric observations and flexible goal specifications. Second, CE-Adapter refines this trace by predicting embodiment-conditioned residual corrections from visual traversability cues, robot identity, and the initial trace. To train the refinement module without costly manual annotation, Cross-Embodiment RRT* (CE-RRT*) converts panoptic segmentation into robot-conditioned traversability cost maps and generates cost-minimizing pixel-space traces. We evaluate CrossTracer on the NaviTrace benchmark, which tests whether a model can generate embodiment-consistent navigation traces from egocentric observations, language instructions, and robot embodiment types. CrossTracer achieves a total score of 45.68, outperforming the strongest evaluated general-purpose baseline, Gemini-2.5-Pro, by 10.01 points, corresponding to a 28.1% relative improvement. Real-world deployment on wheeled and legged robots further shows improved navigation success and execution efficiency.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Human-video and latent-action pretraining, one of 13 directions in Vision-language-action models. Builds on 34 of its 42 references that are in the map. First posted on arXiv 7 Aug 2026.

Builds on

34 of 42 references.

References · 34

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  3. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 3,080
  4. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,887
  5. 2021Masked-attention Mask Transformer for Universal Image Segmentationcited by 2,351
  6. 2017FiLM: Visual Reasoning with a General Conditioning Layercited by 2,333
  7. 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,061
  8. 2024Octo: An Open-Source Generalist Robot Policycited by 1,484
  9. 2023Open X-Embodiment: Robotic Learning Datasets and RT-X Modelscited by 862
  10. 2023NoMaD: Goal Masked Diffusion Policies for Navigation and Explorationcited by 248
  11. 2023ViNT: A Foundation Model for Visual Navigationcited by 222
  12. 2024NaVILA: Legged Robot Vision-Language-Action Model for Navigationcited by 211
22 more
  1. 2022GNM: A General Navigation Model to Drive Any Robotcited by 171
  2. 2024Uni-NaVid: A Video-based Vision-Language-Action Model for Unifying Embodied Navigation Taskscited by 156
  3. 2022ViKiNG: Vision-Based Kilometer-Scale Navigation with Geographic Hintscited by 58
  4. 2025NavDP: Learning Sim-to-Real Navigation Diffusion Policy with Privileged Information Guidancecited by 49
  5. 2023ViPlanner: Visual Semantic Imperative Learning for Local Navigationcited by 46
  6. 2024CoNVOI: Context-aware Navigation using Vision Language Models in Outdoor and Indoor Environmentscited by 43
  7. 2025OmniVLA: An Omni-Modal Vision-Language-Action Model for Robot Navigationcited by 38
  8. 2024LeLaN: Learning A Language-Conditioned Navigation Policy from In-the-Wild Videoscited by 32
  9. 2025CAST: Counterfactual Labels Improve Instruction Following in Vision-Language-Action Modelscited by 29
  10. 2023EVORA: Deep Evidential Traversability Learning for Risk-Aware Off-Road Autonomycited by 21
  11. 2024X-MOBILITY: End-To-End Generalizable Navigation via World Modelingcited by 16
  12. 2025Learning to Drive Anywhere with Model-Based Reannotationcited by 13
  13. 2025NaviTrace: Evaluating Embodied Navigation of Vision-Language Modelscited by 13
  14. 2024NeuPAN: Direct Point Robot Navigation with End-to-End Model-based Learningcited by 11
  15. 2024VL-TGS: Trajectory Generation and Selection using Vision Language Models in Mapless Outdoor Environmentscited by 11
  16. 2023MTG: Mapless Trajectory Generator with Traversability Coverage for Outdoor Navigationcited by 10
  17. 2025CE-Nav: Flow-Guided Reinforcement Refinement for Cross-Embodiment Local Navigationcited by 10
  18. 2025COMPASS: Cross-embodiment Mobility Policy via Residual RL and Skill Synthesiscited by 8
  19. 2025VAMOS: A Hierarchical Vision-Language-Action Model for Capability-Modulated and Steerable Navigationcited by 7
  20. 2025X-Nav: Learning End-to-End Cross-Embodiment Navigation for Mobile Robotscited by 5
  21. 2024Traversability-Aware Legged Navigation by Learning from Real-World Visual Datacited by 4
  22. 2026FLUX: Accelerating Cross-Embodiment Generative Navigation Policies via Rectified Flow and Static-to-Dynamic Learningcited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025EmbodiedDiffusion: End-to-End Traversability-Guided Visual Diffusion for Heterogeneous Robot Navigation
  2. 2026RECAST: Recasting Vision-Language Semantics into an Actionable Cost Map for Robot Navigation
  3. 2026Learning Language-Conditioned Traversability Representations for Adaptive Visual Navigation ↗
  4. 2026CATNAV: Cached Vision-Language Traversability for Efficient Zero-Shot Robot Navigation
  5. 2025VENTURA: Adapting Image Diffusion Models for Unified Task Conditioned Navigation
  6. 2025Rethinking the Embodied Gap in Vision-and-Language Navigation: A Holistic Study of Physical and Visual Disparities
  7. 2026GemNav: Discrete-Token Visual Robot Navigation using a Multimodal Large Language Model
  8. 2026SpaAct: Spatially-Activated Transition Learning with Curriculum Adaptation for Vision-Language Navigation