arXiv subjects

Masato Kobayashi

Publications and source records attributed to Masato Kobayashi.

At least 19 recordsLinked to original sources

arXiv2026

ULOHA: An Underwater Bimanual Robot System for Robot Learning

Underwater visuomotor policy learning has focused primarily on single manipulators, while bimanual imitation learning has been studied largely in air. We present ULOHA, an underwater bimanual robot learning platform that combines custom-designed leader--follower hardware with software extensions to LeRobot, integrating teleoperation, multi-view sensing, demonstration collection, policy training, and autonomous deployment. Real-robot experiments demonstrate a range of coordinated underwater bimanual behaviors, including inter-arm transfer, shared-object manipulation, and buoyancy-driven interception. We evaluate ACT, Diffusion Policy, and the vision--language--action model SmolVLA on the platform. We investigate how learning methods and execution strategies developed for manipulation in air perform underwater, examining bubble disturbances, buoyancy-driven object motion, action-execution horizons, and real-time chunking. A separate single-arm study examines policy transfer between air and water and shows that demonstrations spanning both media support execution in both under the tested conditions. ULOHA provides a unified experimental platform for studying underwater bimanual robot learning under the coupled perceptual and physical effects of underwater environments. Additional material: https://mertcookimg.github.io/uloha/

cs.RO↗

arXiv2026

Bi-MoDe: Bilateral Control-based Imitation Learning via Modifier-Conditioned Decoding for Modulation of Execution Speed and Contact Intensity

Bilateral control-based imitation learning captures both position and force information, making it well suited to contact-rich manipulation. However, existing approaches provide limited means for an operator to specify how a learned task should be executed at inference time, such as slowly or quickly, gently or firmly. We propose Bi-MoDe, a modifier-conditioned decoding framework that injects a constrained latent into every layer of the Transformer action decoder via adaLN-Zero, allowing behavioral directives to directly influence action-chunk generation. We evaluate the method on a real-world whiteboard wiping task with combinations of temporal and physical modifiers. Bi-MoDe improves physical directive following over the action-chunking baseline while maintaining comparable temporal control. An ablation further shows that decoder conditioning and latent-space composition interact, and that their combination is important for accurate physical directive following. Additional material is available at the https://mertcookimg.github.io/bi-mode/

cs.RO↗

arXiv2026

MR-GLi: Mixed Reality-Based Gripper-Linked Overlays for Underwater Robot Arm Teleoperation via Bilateral Control

Visual torque feedback supports underwater bilateral teleoperation, but the benefit of mixed reality (MR) over conventional monitor presentation remains unclear. We present MR-GLi, an MR interface that spatially registers a reaction torque indicator and wrist-camera image to the robot gripper. Twenty participants performed lift and pick-and-place tasks with rigid and compliant objects in a counterbalanced within-subject comparison with a 2D monitor, using identical visual-feedback content and four-channel bilateral control. MR-GLi provided gripper-linked access to visual feedback while maintaining a similar level of torque-regulation performance to the 2D monitor. Subjective evaluation further indicated reduced perceived burden associated with shifting attention between the workspace and visual feedback. These results demonstrate the feasibility of gripper-linked MR overlays for underwater bilateral teleoperation and highlight the importance of considering information access in addition to task performance. Additional material: https://mertcookimg.github.io/mr-gli/

cs.RO↗

arXiv2026

D3DWA: Adaptive Weight and Prediction-Horizon for Dynamic Window Approach via Dueling Double Deep Q-Network

The Dynamic Window Approach (DWA) is widely used for local navigation, but its performance depends strongly on parameters that are typically fixed before navigation. In particular, the appropriate prediction horizon can vary with local free space: longer horizons support efficient motion in open areas, whereas shorter horizons help preserve feasible motions in narrow or cluttered regions. This paper proposes D3DWA, an adaptive DWA framework based on a Dueling Double Deep Q-Network (D3QN), which jointly selects the DWA evaluation weights and prediction horizon from a continuous navigation state at every control step while retaining DWA's trajectory generation and collision checking. In eight simulated environments, including unseen layouts, D3DWA reached every goal. Real-robot experiments further showed that D3DWA completed all three tested configurations, including a constrained case in which the weights-only variant timed out. These results demonstrate the benefit of jointly adapting the evaluation weights and prediction horizon. Additional material is available at https://mertcookimg.github.io/d3dwa/

cs.RO↗

arXiv2026

MRReP: Mixed Reality-based Hand-drawn Reference Path Editing Interface for Mobile Robot Navigation

Autonomous mobile robots operating in human-shared indoor environments often require paths that reflect human spatial intentions, such as avoiding interference with pedestrian flow or maintaining comfortable clearance. However, conventional path planners primarily optimize geometric costs and provide limited support for explicit route specification by human operators. This paper presents MRReP, a Mixed Reality-based interface that enables users to draw a Hand-drawn Reference Path (HRP) directly on the physical floor using hand gestures. The drawn HRP is integrated into the robot navigation stack through a custom Hand-drawn Reference Path Planner, which converts the user-specified point sequence into a global path for autonomous navigation. We evaluated MRReP in a within-subject experiment against a conventional 2D baseline interface. The results demonstrated that MRReP enhanced path specification accuracy, usability, and perceived workload, while enabling more stable path specification in the physical environment. These findings suggest that direct path specification in MR is an effective approach for incorporating human spatial intention into mobile robot navigation. Additional material is available at https://mertcookimg.github.io/mrrep

cs.RO↗

arXiv2026

Bi-HIL: Bilateral Control-Based Multimodal Hierarchical Imitation Learning via Subtask-Level Progress Rate and Keyframe Memory for Long-Horizon Contact-Rich Robotic Manipulation

Long-horizon contact-rich robotic manipulation remains challenging due to partial observability and unstable subtask transitions under contact uncertainty. While hierarchical architectures improve temporal reasoning and bilateral imitation learning enables force-aware control, existing approaches often rely on flat policies that struggle with long-horizon coordination. We propose Bi-HIL, a bilateral control-based multimodal hierarchical imitation learning framework for long-horizon manipulation. Bi-HIL stabilizes hierarchical coordination by integrating keyframe memory with subtask-level progress rate that models phase progression within the active subtask and conditions both high- and low-level policies. We evaluate Bi-HIL on unimanual and bimanual real-robot tasks, demonstrating consistent improvements over flat and ablated variants. The results highlight the importance of explicitly modeling subtask progression together with force-aware control for robust long-horizon manipulation. For additional material, please check: https://mertcookimg.github.io/bi-hil

cs.RO↗

arXiv2026

Bi-AQUA: Bilateral Control-Based Imitation Learning for Underwater Robot Arms via Lighting-Aware Action Chunking with Transformers

Underwater robotic manipulation remains challenging because lighting variation, color attenuation, scattering, and reduced visibility can severely degrade visuomotor policies. We present Bi-AQUA, the first underwater bilateral control-based imitation learning framework for robot arms that explicitly models lighting within the policy. Bi-AQUA integrates transformer-based bilateral action chunking with a hierarchical lighting-aware design composed of a label-free Lighting Encoder, FiLM-based visual feature modulation, and a lighting token for action conditioning. This design enables adaptation to static and dynamically changing underwater illumination while preserving the force-sensitive advantages of bilateral control, which are particularly important in long-horizon and contact-rich manipulation. Real-world experiments on underwater pick-and-place, drawer closing, and peg extraction tasks show that Bi-AQUA outperforms a bilateral baseline without lighting modeling and achieves robust performance under seen, unseen, and changing lighting conditions. These results highlight the importance of combining explicit lighting modeling with force-aware bilateral imitation learning for reliable underwater manipulation. For additional material, please check: https://mertcookimg.github.io/bi-aqua

cs.RO↗

arXiv2026

MRPoS: Mixed Reality-Based Robot Navigation Interface Using Spatial Pointing and Speech with Large Language Model

Recent advancements have made robot navigation more intuitive by transitioning from traditional 2D displays to spatially aware Mixed Reality (MR) systems. However, current MR interfaces often rely on manual "air tap" gestures for goal placement, which can be repetitive and physically demanding, especially for beginners. This paper proposes the Mixed Reality-Based Robot Navigation Interface using Spatial Pointing and Speech (MRPoS). This novel framework replaces complex hand gestures with a natural, multimodal interface combining spatial pointing with Large Language Model (LLM)-based speech interaction. By leveraging both information, the system translates verbal intent into navigation goals visualized by MR technology. Comprehensive experiments comparing MRPoS against conventional gesture-based systems demonstrate that our approach significantly reduces task completion time and workload, providing a more accessible and efficient interface. For additional material, please check: https://mertcookimg.github.io/mrpos

cs.RO↗

arXiv2026

RSK correspondence for King tableaux with Berele insertion

We establish a bijective RSK correspondence of type C for King tableaux with Berele insertion as a reformulation of Sundaram's correspondence (1986). For its $Q$-symbol, we make use of semistandard oscillating tableaux (SSOT), a new object which Lee (2025) introduced. Further, we show hidden duality of Cauchy identity through RSK correspondences of type A and C. Finally, we prove that the generating function of SSOT is symmetric by constructing a new sort of Bender-Knuth involution.

math.CO↗

arXiv2026

Symmetry of the generating function of semistandard oscillating tableaux

H.Choi-D.Kim-S.J.Lee and S.J.Lee introduced a new kind of tableaux, semistandard oscillating tableaux (SSOT), around 2024 in the context of Lusztig $q$-weight multiplicities, KR crystals and King tableaux. In this paper, we study generating function of the SSOTs and its symmetry. First, we extend Gessel's and Assaf-Searles' expansion of a Schur function in terms of fundamental quasi-symmetric functions to our generating function. As a consequence, we show that it is $F$-positive. Further, we improve Sundaram's work on oscillating tableaux by proving that it is symmetric, Schur-positive, and has Saturated Newton polytope.

math.CO↗

arXiv2025

Crystal skeleton polynomials with major index, charge and depth

We introduce a new family of polynomials, crystal skeleton polynomials, to better understand enumeration of standard Young tableaux, quasi-Yamanouchi tableaux and interactions with Gessel's expansion of a Schur function, quasi-crystals and crystal skeletons as Maas-Gariépy introduced in 2023. After developing calculus of those polynomials, we organize thoughts on major index, charge, depth, inversions with RSK correspondence and a bivariate factorial. Also, we revisit the theorem on internal zeros of fake degree polynomials by Billey--Konvalinka--Swanson (2020). These results altogether improve Gessel's expansion.

math.CO↗

arXiv2025

MR-UBi: Mixed Reality-Based Underwater Robot Arm Teleoperation System with Reaction Torque Indicator via Bilateral Control

We present a mixed reality-based underwater robot arm teleoperation system with a reaction torque indicator via bilateral control (MR-UBi). The reaction torque indicator (RTI) overlays a color and length-coded torque bar in the MR-HMD, enabling seamless integration of visual and haptic feedback during underwater robot arm teleoperation. User studies with sixteen participants compared MR-UBi against a bilateral-control baseline. MR-UBi significantly improved grasping-torque control accuracy, increasing the time within the optimal torque range and reducing both low and high grasping torque range during lift and pick-and-place tasks with objects of different stiffness. Subjective evaluations further showed higher usability (SUS) and lower workload (NASA--TLX). Overall, the results confirm that \textit{MR-UBi} enables more stable, accurate, and user-friendly underwater robot-arm teleoperation through the integration of visual and haptic feedback. For additional material, please check: https://mertcookimg.github.io/mr-ubi

cs.RO↗

arXiv2025

Disentangling Static and Dynamic Information for Reducing Static Bias in Action Recognition

Action recognition models rely excessively on static cues rather than dynamic human motion, which is known as static bias. This bias leads to poor performance in real-world applications and zero-shot action recognition. In this paper, we propose a method to reduce static bias by separating temporal dynamic information from static scene information. Our approach uses a statistical independence loss between biased and unbiased streams, combined with a scene prediction loss. Our experiments demonstrate that this method effectively reduces static bias and confirm the importance of scene prediction loss.

cs.CV↗

arXiv2025

Bi-VLA: Bilateral Control-Based Imitation Learning via Vision-Language Fusion for Action Generation

We propose Bilateral Control-Based Imitation Learning via Vision-Language Fusion for Action Generation (Bi-VLA), a novel framework that extends bilateral control-based imitation learning to handle more than one task within a single model. Conventional bilateral control methods exploit joint angle, velocity, torque, and vision for precise manipulation but require task-specific models, limiting their generality. Bi-VLA overcomes this limitation by utilizing robot joint angle, velocity, and torque data from leader-follower bilateral control with visual features and natural language instructions through SigLIP and FiLM-based fusion. We validated Bi-VLA on two task types: one requiring supplementary language cues and another distinguishable solely by vision. Real-robot experiments showed that Bi-VLA successfully interprets vision-language combinations and improves task success rates compared to conventional bilateral control-based imitation learning. Our Bi-VLA addresses the single-task limitation of prior bilateral approaches and provides empirical evidence that combining vision and language significantly enhances versatility. Experimental results validate the effectiveness of Bi-VLA in real-world tasks. For additional material, please visit the website: https://mertcookimg.github.io/bi-vla/

cs.RO↗

arXiv2025

Spatially and Dynamically Extended Molecular Gas in Stephan's Quintet Revealed by ALMA CO(1-0) Total Power Mapping

We present ALMA Total Power CO(1-0) mapping of Stephan's Quintet (SQ), a prototypical compact galaxy group, with a uniform noise level at a spatial scale of ~25 kpc. These observations provide the first complete view of molecular gas across the whole system. Molecular gas is found to spread over a wide area (~120 x 80 kpc), mainly over the two main member galaxies (NGC7318B and 7319), but also in the shocked ridges between these galaxies, the tidal tail, and also in intergalactic regions north of the tail. The total CO(1-0) luminosity is $(2.47\pm0.12)\times10^9~\mathrm{K~km~s^{-1}~pc^2}$, corresponding to a molecular gas mass of $(1.07\pm0.05)\times10^{10}~M_\odot$ assuming the Galactic CO-to-H2 conversion factor. The global star formation efficiency of SQ is estimated at 0.29-0.70 $\mathrm{Gyr^{-1}}$, comparable to or lower than that of nearby star-forming galaxies. Molecular gas spans a velocity range of ~1300 km/s, which can be divided into three components (low, mid, high). The low- and mid-velocity components, linked to NGC7318B and the ridge, show relatively active star formation, whereas the high-velocity component, associated with NGC7319, shows suppressed star formation. Our mapping reveals molecular gas extending ~100 kpc in projection along the inner tail and north of it, containing $(1.64\pm0.08)\times10^9~M_\odot$ (15% of total) with low velocity dispersion (~20 km/s) and ongoing star formation. While previous studies suggested in situ molecular gas formation in the tail, our data suggest an additional contribution from gas stripped from NGC7319.

astro-ph.GA↗

arXiv2025

The 1st International Workshop on Disentangled Representation Learning for Controllable Generation (DRL4Real): Methods and Results

This paper reviews the 1st International Workshop on Disentangled Representation Learning for Controllable Generation (DRL4Real), held in conjunction with ICCV 2025. The workshop aimed to bridge the gap between the theoretical promise of Disentangled Representation Learning (DRL) and its application in realistic scenarios, moving beyond synthetic benchmarks. DRL4Real focused on evaluating DRL methods in practical applications such as controllable generation, exploring advancements in model robustness, interpretability, and generalization. The workshop accepted 9 papers covering a broad range of topics, including the integration of novel inductive biases (e.g., language), the application of diffusion models to DRL, 3D-aware disentanglement, and the expansion of DRL into specialized domains like autonomous driving and EEG analysis. This summary details the workshop's objectives, the themes of the accepted papers, and provides an overview of the methodologies proposed by the authors.

cs.LG↗

arXiv2025

MRHaD: Mixed Reality-based Hand-Drawn Map Editing Interface for Mobile Robot Navigation

Mobile robot navigation systems are increasingly relied upon in dynamic and complex environments, yet they often struggle with map inaccuracies and the resulting inefficient path planning. This paper presents MRHaD, a Mixed Reality-based Hand-drawn Map Editing Interface that enables intuitive, real-time map modifications through natural hand gestures. By integrating the MR head-mounted display with the robotic navigation system, operators can directly create hand-drawn restricted zones (HRZ), thereby bridging the gap between 2D map representations and the real-world environment. Comparative experiments against conventional 2D editing methods demonstrate that MRHaD significantly improves editing efficiency, map accuracy, and overall usability, contributing to safer and more efficient mobile robot operations. The proposed approach provides a robust technical foundation for advancing human-robot collaboration and establishing innovative interaction models that enhance the hybrid future of robotics and human society. For additional material, please check: https://mertcookimg.github.io/mrhad/

cs.RO↗

arXiv2025

Bi-LAT: Bilateral Control-Based Imitation Learning via Natural Language and Action Chunking with Transformers

We present Bi-LAT, a novel imitation learning framework that unifies bilateral control with natural language processing to achieve precise force modulation in robotic manipulation. Bi-LAT leverages joint position, velocity, and torque data from leader-follower teleoperation while also integrating visual and linguistic cues to dynamically adjust applied force. By encoding human instructions such as "softly grasp the cup" or "strongly twist the sponge" through a multimodal Transformer-based model, Bi-LAT learns to distinguish nuanced force requirements in real-world tasks. We demonstrate Bi-LAT's performance in (1) unimanual cup-stacking scenario where the robot accurately modulates grasp force based on language commands, and (2) bimanual sponge-twisting task that requires coordinated force control. Experimental results show that Bi-LAT effectively reproduces the instructed force levels, particularly when incorporating SigLIP among tested language encoders. Our findings demonstrate the potential of integrating natural language cues into imitation learning, paving the way for more intuitive and adaptive human-robot interaction. For additional material, please visit: https://mertcookimg.github.io/bi-lat/

cs.RO↗