Yuki Fujimura
Portrait of Yuki Fujimura

Computer Vision · Machine Learning

Yuki Fujimura

Associate Professor

Optical Media Interface Laboratory
Nara Institute of Science and Technology (NAIST)

Yuki Fujimura received the PhD in informatics at Kyoto University in 2021. His research interests include physics-based vision and machine learning.

fujimura.yuki [at] is.naist.jp

Selected work

Research Projects

Feed-forward 3D Gaussian splatting for unconstrained images

2026 · arXiv

Feed-forward 3D Gaussian splatting for unconstrained images

Yuki Fujimura, Takahiro Kushida, Kitano Kazuya, Takuya Funatomi, and Yasuhiro Mukaigawa, WildSplatter: Feed-forward 3D Gaussian Splatting with Appearance Control from Unconstrained Images, arXiv:2604.21182, 2026.

Project ↗ Paper ↗
Multi-view camera images and a reconstructed 3D camera model

2025 · arXiv

Pose-free feed-forward 3D Gaussian splatting

Yuki Fujimura, Takahiro Kushida, Kitano Kazuya, Takuya Funatomi, and Yasuhiro Mukaigawa, UFV-Splatter: Pose-Free Feed-Forward 3D Gaussian Splatting Adapted to Unfavorable Views, arXiv:2507.22342, 2025.

Project ↗ Paper ↗
Camera defocus model and neural depth estimation diagram

2024 · IJCV

Deep depth from focus/defocus with a camera model

Yuki Fujimura, Masaaki Iiyama, Takuya Funatomi, and Yasuhiro Mukaigawa, Deep Depth from Focal Stack with Defocus Model for Camera-Setting Invariance, IJCV, 2024.

Paper ↗ Code ↗ arXiv ↗ Press ↗
Non-line-of-sight imaging setup and reconstructed sculpture

2023 · ICCV

Neural implicit surface in non-line-of-sight imaging

Yuki Fujimura, Takahiro Kushida, Takuya Funatomi, and Yasuhiro Mukaigawa, NLOS-NeuS: Non-line-of-sight Neural Implicit Surface, ICCV, 2023.

Project ↗ Paper ↗ arXiv ↗
Dehazing cost volume for multi-view stereo diagram

2020 · ACCV Oral

Deep multi-view stereo in foggy scenes

Yuki Fujimura, Motoharu Sonogashira, and Masaaki Iiyama, Dehazing Cost Volume for Deep Multi-view Stereo in Scattering Media, ACCV (Oral), 2020.

Paper ↗ Code ↗
Photometric stereo inputs, normal maps, and recovered shapes

2018–2020 · CVPR Oral / TPAMI

Photometric stereo in participating media

Yuki Fujimura, Masaaki Iiyama, Atsushi Hashimoto, and Michihiko Minoh, Photometric Stereo in Participating Media Considering Shape-Dependent Forward Scatter, CVPR (Oral), 2018; Photometric Stereo in Participating Media Using an Analytical Solution for Shape-Dependent Forward Scatter, TPAMI, 2020.

CVPR ↗ TPAMI ↗

Recognition

Awards

Research support

Grants