Hiroyuki Sato

Prof. Hiroyuki Sato

The University of Tokyo, Tokyo, Japan
orcid

Research Focus & Metrics

Academic Career Length

32Years (1995-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 July 2024 in ACM Transactions on Graphics
Verified
Journal Article
Open access
ZWZhennan Wu
Zhennan Wu
YLYang Li
Yang Li
HYHan Yan
Han Yan
TSTaizhang Shang
Taizhang Shang
WSWeixuan Sun
Weixuan Sun
PJPan Ji
Pan Ji
We present BlockFusion, a diffusion-based model that generates 3D scenes as unit blocks and seamlessly incorporates new blocks to extend the scene. BlockFusion is trained using datasets of 3D blocks that are randomly cropped from complete 3D scene meshes. Through per-block fitting, all training blocks are converted into the hybrid neural fields: with a tri-plane containing the geometry features, followed by a Multi-layer Perceptron (MLP) for decoding the signed distance values. A variational auto-encoder is employed to compress the tri-planes into the latent tri-plane space, on which the denoising diffusion process is performed. Diffusion applied to the latent representations allows for high-quality and diverse 3D scene generation.To expand a scene during generation, one needs only to append empty blocks to overlap with the current scene and extrapolate existing latent tri-planes to populate new blocks. The extrapolation is done by conditioning the generation process with the feature samples from the overlapping tri-planes during the denoising iterations. Latent tri-plane extrapolation produces semantically and geometrically meaningful transitions that harmoniously blend with the existing scene. A 2D layout conditioning mechanism is used to control the placement and arrangement of scene elements. Experimental results indicate that BlockFusion is capable of generating diverse, geometrically consistent and unbounded large 3D scenes with unprecedented high-quality shapes in both indoor and outdoor scenarios.

7 October 2022 in Journal of Network and Systems Management
Verified
Journal Article
Closed access
PDF file
YDYepeng Ding
Yepeng Ding
Hiroyuki SatoHiroyuki Sato
Hiroyuki Sato

25 May 2022 in IEEE Access
Verified
Journal Article
Open access
GLGen Li
Gen Li
Hiroyuki SatoHiroyuki Sato
Hiroyuki Sato
By virtue of the stability of signatures and the high difficulty of imitation, handwriting signatures, as an important behavioral biometric trait, have been broadly adopted for authorization and identity verification. The emergence of consumer-level wrist-worn devices incorporating rich sensors has profoundly changed human-machine interactions, enabling new signature observation method. In this study, we investigate the feasibility of authenticating users by sensing hand motions of signing in air using fingers. Each signature is represented by the readings of the gyroscope and accelerometer which are compensated by the device attitude readings. A recurrent neural network-based algorithm is proposed to characterize signatures and accurately determine whether a signature is from the claimed genuine user or an imposter. We empirically investigate 22 participants by recording their in-air signing gestures using smartwatch motion sensors. The verification shows that despite the inevitable variability of repeating genuine signature drawing, forged signatures tend to show more dissimilarity than variability. The high-precision experimental result (i.e., equal error rate of 0.83%) against insider adversaries not only demonstrates the effectiveness of our proposed approach but also indicates the feasibility of a more user-friendly signature authentication method by signing their names in the air. Moreover, we investigate the impact of the properties of motion sensory data on signature authentication. In addition, we include more details of the experiments, validation of the proposed pre-processing method, and analysis of the circumvention as one of the desirable properties of biometrics of signing motions by measuring the skill of forgery.

16 September 2025 in Proceedings of the International Conference on Automated Planning and Scheduling
Verified
Journal Article
Open access
PDF file
JYJunwei Yu
Junwei Yu
YDYepeng Ding
Yepeng Ding
Hiroyuki SatoHiroyuki Sato
Hiroyuki Sato
The emergence of Large Language Models (LLMs) in Multi-Agent Systems (MAS) has opened new possibilities for artificial intelligence, yet current implementations face significant challenges in resource management, task coordination, and system efficiency. While existing frameworks demonstrate the potential of LLM-based agents in collaborative problem-solving, they often lack sophisticated mechanisms for parallel execution and dynamic task management. This paper introduces DynTaskMAS, a novel framework that orchestrates asynchronous and parallel operations in LLM-based MAS through dynamic task graphs. The framework features four key innovations: (1) a Dynamic Task Graph Generator that intelligently decomposes complex tasks while maintaining logical dependencies, (2) an Asynchronous Parallel Execution Engine that optimizes resource utilization through efficient task scheduling, (3) a Semantic-Aware Context Management System that enables efficient information sharing among agents, and (4) an Adaptive Workflow Manager that dynamically optimizes system performance. Experimental evaluations demonstrate that DynTaskMAS achieves significant improvements over traditional approaches: a 21-33% reduction in execution time across task complexities (with higher gains for more complex tasks), a 35.4% improvement in resource utilization (from 65% to 88%), and near-linear throughput scaling up to 16 concurrent agents (3.47× improvement for 4× agents). Our framework establishes a foundation for building scalable, high-performance LLM-based multi-agent systems capable of handling complex, dynamic tasks efficiently.

1 April 2002 in International Journal of Hematology
Verified
Journal Article
Closed access
PDF file
ATAkira Tomonari
Akira Tomonari
TITohru Iseki
Tohru Iseki
JOJun Ooi
Jun Ooi
HNHitomi Nagayama
Hitomi Nagayama
Hiroyuki SatoHiroyuki Sato
Hiroyuki Sato
SAShigetaka Asano
Shigetaka Asano

14 April 2022 in Journal of Alloys and Compounds
Verified
Journal Article
Closed access
TMTakahiro Mineta
Takahiro Mineta
HSHaruto Saijo
Haruto Saijo
Hiroyuki SatoHiroyuki Sato
Hiroyuki Sato