Yung-Hsu (Roy) Yang (楊詠旭)

I am a Ph.D. student at CVG ETH Zürich supervised by Prof. Marc Pollefeys and Prof. Hermann Blum. My research interests include scene understanding and 3D Object Detection and Tracking. Starting June 2026, I work as a student researcher at Google Zürich, hosted by Christina Tsalicoglou, Erik Sandström, and Federico Tombari.

I obtained my M.Sc. and B.Sc. degrees in Electrical Engineering department at National Tsing Hua University supervised by Prof. Min Sun. Previously, I worked with Dr. Samuel Rota Bulò and Dr. Peter Kontschieder in dense prediction tasks.

Email  /  Scholar  /  Github  /  Linkedin  /  X  /  Bluesky  /  Huggingface  /  CV

profile photo

Research
LeAD-M3D: Leveraging Asymmetric Distillation for Real-time Monocular 3D Detection
Johannes Meier,  Jonathan Michel,  Oussema Dhaouadi,  Yung-Hsu Yang,  Christoph Reich,  Zuria Bauer,  Stefan Roth,  Marc Pollefeys,  Jacques Kaiser,  Daniel Cremers
ECCV, 2026
arXiv / Code
DVPSFormer: Efficient Online Depth-aware Video Panoptic Segmentation for Autonomous Driving
Yung-Hsu Yang,  Luigi Piccinelli,  Siyuan Li,  Mattia Segu,  Lei Ke,  Martin Danelljan,  Yuqian Fu,  Zuria Bauer,  Fisher Yu,  Hermann Blum,  Marc Pollefeys
arXiv, 2026
arXiv / Code
3D-MOOD: Lifting 2D to 3D for Monocular Open-Set Object Detection
Yung-Hsu Yang,  Luigi Piccinelli,  Mattia Segu,  Siyuan Li,  Rui Huang,  Yuqian Fu,  Marc Pollefeys,  Hermann Blum,  Zuria Bauer
ICCV, 2025
arXiv / Code / Demo
FlowR: Flowing from Sparse to Dense 3D Reconstructions
Tobias Fischer,  Samuel Rota Bulò,  Yung-Hsu Yang,  Nikhil Varma Keetha,  Lorenzo Porzi,  Norman Müller,  Katja Schwarz,  Jonathon Luiten,  Marc Pollefeys,  Peter Kontschieder
ICCV, 2025 (Highlight)
arXiv / Code
UniK3D: Universal Camera Monocular 3D Estimation
Luigi Piccinelli,  Christos Sakaridis,  Mattia Segu,  Yung-Hsu Yang,  Siyuan Li,  Wim Abbeloos,  Luc Van Gool
CVPR, 2025
arXiv / Code / Demo
UniDepthV2: Universal Monocular Metric Depth Estimation Made Simpler
Luigi Piccinelli,  Christos Sakaridis,  Yung-Hsu Yang,  Mattia Segu,  Siyuan Li,  Wim Abbeloos,  Luc Van Gool
TPAMI, 2025
arXiv / Code
Samba: Synchronized Set-of-Sequences Modeling for End-to-end Multiple Object Tracking
Mattia Segu,  Luigi Piccinelli,  Siyuan Li,  Yung-Hsu Yang,  Bernt Schiele,  Luc Van Gool
ICLR, 2025 (Spotlight)
arXiv / Code
SLAck: Semantic, Location, and Appearance Aware Open-Vocabulary Tracking
Siyuan Li,  Lei Ke,  Yung-Hsu Yang,  Luigi Piccinelli,  Mattia Segu,  Martin Danelljan,  Luc Van Gool
ECCV, 2024
arXiv / Code
UniDepth: Universal Monocular Metric Depth Estimation
Luigi Piccinelli,  Yung-Hsu Yang,  Christos Sakaridis,  Mattia Segu,  Siyuan Li,  Luc Van Gool,  Fisher Yu
CVPR, 2024 (Highlight)
arXiv / Code / Video
CR3DT: Camera-RADAR Fusion for 3D Detection and Tracking
Nicolas Baumann*,  Michael Baumgartner*,  Edoardo Ghignone*,  Jonas Kühne*,  Tobias Fischer,  Yung-Hsu Yang,  Marc Pollefeys,  Michele Magno
IROS, 2024
arXiv / Code
Dense Prediction with Attentive Feature Aggregation
Yung-Hsu Yang,  Thomas E. Huang,  Min Sun,  Samuel Rota Bulò,  Peter Kontschieder,  Fisher Yu
WACV, 2023
arXiv / Code / Video
CC-3DT: Panoramic 3D Object Tracking via Cross-Camera Fusion
Tobias Fischer*,  Yung-Hsu Yang*,  Suryansh Kumar,  Min Sun,  Fisher Yu
CoRL, 2022
arXiv / Code / Video
Monocular Quasi-Dense 3D Object Tracking
Hou-Ning Hu,  Yung-Hsu Yang,  Tobias Fischer,  Trevor Darrell,  Fisher Yu,  Min Sun
TPAMI, 2022
arXiv / Code / Video

Project
Vis4D: A modular library for 4D scene understanding
Yung-Hsu Yang*,  Tobias Fischer*,  Thomas E. Huang*,  Tao Sun,  René Zurbrügg,  Fisher Yu

Academic Services

  • Conference Reviewer
    • CVPR: 2023, 2024, 2025, 2026
    • ICCV: 2025
    • ECCV: 2024, 2026
    • NeurIPS: 2023, 2024, 2025
    • ICLR: 2024, 2025
    • ICML: 2025
    • AAAI: 2026
    • ACM MM: 2024
    • 3DV: 2025, 2026
    • WACV: 2026
    • BMCV: 2026
  • Journal Reviewer
    • TPAMI
    • RA-L
    • TIP
    • IJCV

Awesome website template