Institution profile

National Institute of Advanced Industrial Science and Technology

Academic institutionasia · jp
Official website
Research library255linked papers
Opportunities0open roles
Selected work

Representative Papers

Replacing Labeled Real-image Datasets with Auto-generated Contours

Jun 01, 2022Computer Vision and Pattern Recognition

Pretraining vision transformers (ViTs) typically relies on large-scale real-image datasets, raising concerns regarding data privacy, environmental cost, and annotation effort. Method: We propose Formula-Driven Supervised Learning (FDSL), the first framework enabling ViT pretraining exclusively on synthetically generated contour images—without real images, human annotations, or self-supervision. Contours are procedurally generated via mathematical formulas, yielding controllable complexity, zero bias, zero cost, and zero privacy risk. Contribution/Results: We demonstrate that contour structures alone encode sufficient semantic information for effective representation learning, and that moderately increasing pretraining task difficulty improves transfer performance. A ViT-Base pretrained via FDSL achieves 82.7% top-1 accuracy on ImageNet-1K fine-tuning—surpassing the ImageNet-21K baseline (81.8%). This establishes that purely synthetic contour data can match—or even exceed—the efficacy of large-scale real-image pretraining, opening a new pathway toward green, trustworthy, and interpretable vision foundation models.

32 citations2 influentialRead paper

Humanoid Loco-Manipulation Planning Based on Graph Search and Reachability Maps

Apr 01, 2021IEEE Robotics and Automation Letters

This work addresses the challenge of locomotion-manipulation co-planning for autonomous object transport by humanoid robots. We propose a graph-search-based sequential motion planning framework. Our key contributions are: (1) a novel state transition model enabling flexible coupling between gait and grasping actions; and (2) the first integration of real-time relocalization of reachability maps into the graph search process, enabling efficient online update and switching of dynamic reachable regions under robot–object cooperative motion. The method unifies kinematic modeling, multimodal motion planning, and real-time map construction. Evaluated in simulation and on a physical humanoid platform, it successfully accomplishes complex tasks—including drum rolling and mid-transport regrasping—demonstrating full autonomy, task generality, and real-time performance with millisecond-scale replanning.

27 citationsRead paper

Centroidal Trajectory Generation and Stabilization Based on Preview Control for Humanoid Multi-Contact Motion

Jul 01, 2022IEEE Robotics and Automation Letters

This work addresses the challenge of online centroidal momentum (CoM) trajectory generation and stability control for humanoid robots executing multi-contact dynamic motions in complex environments. We propose a lightweight hierarchical control framework based on preview control, replacing computationally intensive full-horizon constrained model predictive control with a low-complexity preview controller. The method integrates CoM dynamics modeling, state-feedback correction, and optimized contact wrench distribution to explicitly satisfy multi-contact constraints and stability requirements—while ensuring real-time performance (millisecond-level computation). Simulation results demonstrate significant improvements in robustness and trajectory tracking accuracy during representative dynamic tasks, including stepping and support-phase transitions. The approach establishes an efficient and practical paradigm for high-dynamic, multi-contact motion control of humanoid robots.

19 citations1 influentialRead paper

Humanoid Loco-Manipulations Pattern Generation and Stabilization Control

Jul 01, 2021IEEE Robotics and Automation Letters

Humanoid robots face significant challenges in maintaining dynamic balance and coordinating locomotion with manipulation when handling objects during walking, due to persistent or intermittent external contact forces—distinct from ground reaction forces. This paper proposes a dynamics-consistent locomotion-manipulation (locomanipulation) control framework. First, it embeds manipulation force reference trajectories directly into center-of-mass (CoM) motion planning; second, it designs a real-time feedback stabilizer based on external force tracking error. By integrating the linear inverted pendulum model with the divergent component of motion (DCM), the framework enables external-force-aware trajectory generation and robust gait adaptation. Evaluated in simulation and on a physical humanoid platform, the method significantly improves dynamic stability and trajectory tracking accuracy under complex manipulation tasks. It establishes a scalable theoretical and practical paradigm for unified locomotion-manipulation control.

19 citationsRead paper

Boosting Offline Optimizers with Surrogate Sensitivity

Mar 06, 2025International Conference on Machine Learning

Offline optimization of expensive black-box functions in materials engineering suffers from poor robustness due to the high sensitivity of surrogate models to parameter perturbations. Method: We propose, for the first time, an optimizable surrogate sensitivity metric and design a sensitivity-aware regularization method orthogonal to existing frameworks. This approach integrates gradient-based sensitivity analysis with deep-learning-based surrogate modeling and is compatible with mainstream paradigms such as offline Bayesian optimization. Contribution/Results: Evaluated on multiple materials design benchmarks, our method significantly improves optimization success rate (average gain of +23.6%) and solution quality (objective value improvement up to 17.4%). Empirical results demonstrate that explicit sensitivity control delivers critical performance gains for offline optimization of expensive black-box functions in materials engineering.

2 citations1 influentialRead paper
Recent publications

Latest Papers