aiqbench analysis
This research demonstrates a compact optical-flow representation that enables vision-based tactile sensors to classify object rotation across diverse gravity domains, including Earth, Mars, Moon, and orbit. By training a single model on multi-gravity data, the approach mitigates significant domain shifts caused by gravity variations, achieving high accuracy while reducing feature count and inference latency for resource-constrained space robots. These findings are critical for advancing autonomous robotic manipulation in extraterrestrial environments where local gravity conditions differ from Earth norms.
A Minimal Optical-Flow Representation for Vision-Based Tactile Rotation Classification in Robotic Manipulation Across Gravity Domains
arXiv:2610.12073v1 Announce Type: new Abstract: Vision-based tactile sensors provide rich contact information, but processing high-resolution images can be costly for resource-constrained platforms such as space robots. This work investigates whether a compact representation of tactile motion can classify object rotation across different gravity conditions. Dense optical flow from a simulated GelSight Mini is aggregated over a 7x9 grid into 126 features and used to classify the direction of load-induced rotation under Earth, Mars, Moon, and orbital gravity. Gravity causes a small but significant shift in these features, accounting for 1.6% of their variance (R2 = 0.016). Despite its small magnitude, this shift affects models trained only on Earth data: XGBoost accuracy decreases from 94.4% on Earth to 75.9% in orbit. In contrast, a single model trained across all four gravity domains achieves 96.3% overall accuracy and 95.1%-97.0% across individual domains, without using gravity as an input. The representation can also be reduced to 40 features while retaining 95.7% accuracy, with XGBoost requiring only 0.14 ms per inference. These findings show that Earth-gravity performance alone is insufficient to establish the transferability of tactile perception for space robotic manipulation, highlighting the need to account for gravity-induced domain shifts during training and validation.
Abstract:Vision-based tactile sensors provide rich contact information, but processing high-resolution images can be costly for resource-constrained platforms such as space robots. This work investigates whether a compact representation of tactile motion can classify object rotation across different gravity conditions. Dense optical flow from a simulated GelSight Mini is aggregated over a 7x9 grid into 126 features and used to classify the direction of load-induced rotation under Earth, Mars, Moon, and orbital gravity. Gravity causes a small but significant shift in these features, accounting for 1.6% of their variance (R2 = 0.016). Despite its small magnitude, this shift affects models trained only on Earth data: XGBoost accuracy decreases from 94.4% on Earth to 75.9% in orbit. In contrast, a single model trained across all four gravity domains achieves 96.3% overall accuracy and 95.1%-97.0% across individual domains, without using gravity as an input. The representation can also be reduced to 40 features while retaining 95.7% accuracy, with XGBoost requiring only 0.14 ms per inference. These findings show that Earth-gravity performance alone is insufficient to establish the transferability of tactile perception for space robotic manipulation, highlighting the need to account for gravity-induced domain shifts during training and validation.
Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
arXiv:2610.12073 [cs.RO]
(or arXiv:2610.12073v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2610.12073 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 294 words
Related coverage
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Latent Space · Oct 13, 2026
From Optical Flow to Safety: Visual Safety Certification via Scale-Free Relative Motion
arXiv · Oct 16, 2026
Links #5: 2026/07
LessWrong · Jul 31, 2026
Game Design Creativity evals
LessWrong · Oct 12, 2026