CORTEXA
← Browse
arxivcs.ROcs.CV2026-06-29

SIR: Structured Image Representations for Explainable Robot Learning

Paul Mattes, Jan Schwab, Jens Bosch, Nils Blank, Maximilian Xiling Li, Minh-Trung Tang, Moritz Haberland, Rudolf Lioutikov

Existing robot policies based on learned visual embeddings lack explicit structure and are sensitive to visual distractions. Thus, the representations that drive their behaviour are often opaque, making their decision-making process difficult to interpret. To address this, we introduce Structured Image Representations (SIR), a method that leverages Scene Graphs (SGs) as an intermediate representation for robot policy learning. Our approach first constructs a fully connected graph, using image-derived features as initial node representations. Then, a module learns to sparsify this graph end-to-end, creating a task-relevant sub-graph that is passed to the action generation model. This process makes our model intrinsically explainable. Evaluations on RoboCasa show that our sparse graph policies outperform image-based baselines on average with 19.5% vs 14.81% success rate. Most importantly, we show that the learned sparse graphs are a powerful tool for model analysis. By analysing when the model's sub-graph deviates from human expectation, such as by including distractor nodes or omitting key objects, we successfully uncover dataset biases, including spurious correlations and positional biases. https://github.com/intuitive-robots/SIR_Model

View free PDFSource page

Related papers

arxivcs.CVcs.LGcs.RO2026-07-15

Depth-Regularized JEPA World Models Learn More Transferable Representations from Real Outdoor Robot Data

Usman M. Khan

World models, especially based on JEPA architectures, have been shown to learn robust dynamics of various environments. However, learning from visually complex real-world data remains a challenge, especially in unpredictable outdoor environments. We introduce depth as a geometric…

View free PDFSource page
arxivcs.ROcs.AIcs.CV2026-07-15

Semantic Anchoring for Robotic Action Representations

Yuan Xu, Youheng Shi, Chengyang Li, Wentao Zhu, Yizhou Wang

Vision-Language-Action (VLA) models inherit rich semantic representations from pretrained Vision-Language Models, yet fine-tuning on limited robot demonstrations degrades this structure and undermines generalization. A fundamental question therefore arises: what constitutes a goo…

View free PDFSource page
arxivcs.ROcs.CV2026-07-09

SplatCtrl: Perception-Action Coupling via Gaussian Scene Representations and Reactive Robot Control

Siddarth Jain, Ho Jin Choi

Robotic manipulators excel in structured environments but face substantial challenges in unstructured and dynamic settings. This paper presents SplatCtrl, a unified framework for real-time scene reconstruction and reactive robot motion generation to enable collision-free robotic…

View free PDFSource page
arxivcs.CVcs.RO2026-06-25

OctoSense: Self-Supervised Learning for Multimodal Robot Perception

Anthony Bisulco, Jeremy Wang, Kostas Daniilidis, Randall Balestriero, Pratik Chaudhari

We present OctoSense, an open-source sensor platform with stereo RGB and event cameras, LiDAR, a thermal camera, an inertial measurement unit, RTK-corrected global positioning system, and proprioception (CAN bus data from a car, and joint angles for a quadruped robot). The eponym…

View free PDFSource page
arxivcs.AIcs.CVcs.RO2026-07-16

Action QFormer: Structured Representation Shaping under Action Supervision in Vision-Language-Action Models

Yufeng Ji, Wenhao Tang, Haoyi Niu, Koushil Sreenath, Yi Wu, Zhongyu Li

Action supervision in vision-language-action (VLA) models is often treated as a downstream objective for learning action prediction. In this paper, we study it instead as a force that shapes inherited multimodal representations. We show that this shaping has a dual effect: it is…

View free PDFSource page
arxivcs.CVcs.AIcs.RO2026-07-24

SM4RT: Learning Structured Motion Geometry for 4D Reconstruction

Shing Ho J. Lin, Wenzhao Zheng, Dong Zhuo, Yuqi Wu, Jie Zhou, Jiwen Lu

Geometry Foundation Models (GFMs) have substantially advanced monocular 3D reconstruction, yet extending this capability to 4D dynamic understanding remains a fundamental challenge. Most existing motion perception methods (e.g., sparse tracking, dense point-wise flow) treat motio…

View free PDFSource page