Attention-Based Surface Representation Learning for Robot State Prediction and Open-Ended Surface Classification

Attention-Based Surface Representation Learning for Robot State Prediction and Open-Ended Surface Classification

|Authors: Oleg Kushnarev, Alexander Belyaev

|Venue: arXiv preprint | Year: 2026

|arXiv: 2610.04240

Abstract

For ground robots operating in outdoor environments, understanding the properties of the underlying terrain is essential for ensuring reliable operation. In most perception-based studies, this problem is formulated as categorical classification with a fixed number of classes defined during training. We propose an approach that enables new surface classes to be added as trainable vectors, which can subsequently be used to address higher-level tasks. By employing a learning paradigm based on predicting the robot’s next state in time and using attention blocks, we improved classification accuracy to 98.56% on the Belyaev-Kushnarev dataset and 94.8% on BorealTC.

Key Contributions

• For ground robots operating in outdoor environments, understanding the properties of the underlyi…
• In most perception-based studies, this problem is formulated as categorical classification with a…
• We propose an approach that enables new surface classes to be added as trainable vectors, which c…
• By employing a learning paradigm based on predicting the robot’s next state in time and using att…

Topics

• vision
• reinforcement-learning
• benchmark

Code & Data

• GitHub repository: https://github.com/okushnarev/transformer-cls

BibTeX

@article{Kushnarev2026_261004240,
  title     = {Attention-Based Surface Representation Learning for Robot State Prediction and Open-Ended Surface Classification},
  author    = {Oleg Kushnarev and Alexander Belyaev},
  year      = {2026},
  eprint    = {2610.04240},
  archivePrefix = {arXiv},
  primaryClass  = {cs.RO},
  url       = {https://arxiv.org/abs/2610.04240}
}