cv

Doctoral Researcher

University of Tsukuba, CGG Lab

Research on digital human reconstruction and editing using 3D Gaussian Splatting, NeRF, and geometry-aware learning methods.

• Design and implement multi-view training and inference pipelines in PyTorch.
• Study animatable avatars, outfit personalization, and mesh-based deformation models.
• Prototype Point Transformer-based vertex deformation and Gaussian-on-mesh animation pipelines.
• Build tooling for Linux GPU-cluster experimentation and reproducible research workflows.

Master's Researcher

University of Tsukuba, CGG Lab

Research on sparse-view digital human reconstruction and generalizable human NeRF.

• Developed a generalizable human avatar method for sparse input views.
• Worked on differentiable rendering, human priors, and multi-view geometry.

Research Contributor in Face Rendering

Collaborative Research Project

Product-oriented collaborative research on face rendering and inverse rendering.

• Contributed practical modeling and analysis pipelines for facial appearance research.

Ph.D.

University of Tsukuba

Computer Science

M.S.

University of Tsukuba

Computer Science

B.S.

Jiangnan University

Information and Computational Science

AvatarMix: Identity-Preserving Cross-Avatar Composition for Outfit Personalization

CVPR 2026 Findings

Identity-preserving cross-avatar composition for outfit personalization with digital human avatars.

EG-HumanNeRF: Efficient Generalizable Human NeRF Utilizing Human Prior for Sparse View

Computational Visual Media

A generalizable human NeRF method for sparse-view reconstruction using human priors.