profile photo

Kaede Shiohara

I am a postdoctoral researcher at Computer Vision and Media Lab at the University of Tokyo.
I am mainly working on Spatial Artificial Intelligence (SpatialAI), e.g., spatial layout generation, 3D/4D reconstruction and generation.

News
  • 2026.02 A first author and a co-author papers were accepted to CVPR26 Main Track.
    A first author paper was suggested to CVPR26 Findings Track.
  • 2025.11 A co-author paper was accepted to WACV26.
  • 2025.04 I started to work as a postdoctral researcher at the University of Tokyo.
    A co-author paper was accepted to FG26.
  • 2025.03 I won the Department Chair's Ph.D. Thesis Award.
  • 2024.12 I won IEEE Computer Society Japan Chapter Tokyo/Japan Joint Local Award.
  • 2024.08 A co-first author ECCV24 paper was selected as an oral presentation.
  • 2024.07 A co-first author paper was accepted to ECCV24.
  • 2024.06 A co-first author paper was accepted to ICIP24.
    I gave an invited talk at SSII24.
  • 2024.02 A first author paper was accepted to CVPR24.
  • 2023.12 I was accepted to 覚醒.
  • 2023.09 I gave an invited talk at VC23.
  • 2023.08 I started a colaborative research project with MPII.
  • 2023.07 A first author paper was accepted to ICCV23.
  • 2022.09 I was accepted to 学振DC2.
  • 2023.07 I started to work at CyberAgent as a research intern.
  • 2023.03 I won テレコムシステム技術賞.
  • 2022.10 I gave an invited talk at VC22.
  • 2022.07 I gave an invited talk at MIRU22.
  • 2022.06 I won 原島博学術奨励賞.
  • 2022.04 I started my Ph.D. life at the University of Tokyo.
    A first author CVPR22 paper was selected as an oral presentation.
  • 2022.03 A first author paper was accepted to CVPR22.
Selected Publications

Unified Vector Floorplan Generation via Markup Representation

Kaede Shiohara, Toshihiko Yamasaki

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2026 Main Track

ExposeAnyone: Personalized Audio-to-Expression Diffusion Models are Robust Zero-Shot Face Forgery Detectors

Kaede Shiohara, Toshihiko Yamasaki, Vladislav Golyanik

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2026 Findings Track

[paper] [dataset] [project]
PetFace

PetFace: A Large-Scale Dataset and Benchmark for Animal Identification

Risa Shinoda*, Kaede Shiohara* (*Equal contribution)

European Conference on Computer Vision (ECCV) 2024 (Oral, Strong Accept x3)

[paper] [code] [project]
Face2Diffusion

Face2Diffusion for Fast and Editable Face Personalization

Kaede Shiohara, Toshihiko Yamasaki

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2024

[paper] [code] [project]
BlendFace

BlendFace: Re-designing Identity Encoders for Face-Swapping

Kaede Shiohara, Xingchao Yang, Takafumi Taketomi

IEEE/CVF International Conference on Computer Vision (ICCV) 2023

[paper] [code] [project]
Self-Blended Images

Detecting Deepfakes with Self-Blended Images

Kaede Shiohara, Toshihiko Yamasaki

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2022 (Oral)

[paper] [code]
Education

Ph.D. at the University of Tokyo

Graduate School of Information Science and Technology
Department of Information and Communication Engineering
Supervisor: Prof. Toshihiko Yamasaki
2022.4 - 2025.3

M.S. at the University of Tokyo

Graduate School of Information Science and Technology
Department of Information and Communication Engineering
Supervisor: Prof. Toshihiko Yamasaki
2020.4 - 2022.3

B.S. at Osaka City University

Under Graduate School of Engineering
Department of Electrical and Information Engineering
Supervisor: Prof. Kazunori Hayashi
2015.4 - 2019.3