Doctoral student at ETH Zürich
Bio
I am a second-year PhD student supervised by Prof. Dr. Otmar Hilliges. I obtained my Computer Science Master's degree at ETH Zürich in 2021. My research interests lie in human-centric computer vision ranging from basic understandings of human-centric videos, such as human recognition and pose estimation, to more complicated generative tasks, such as motion prediction and full-body generation. Recently, I focus on incorporating user control signals, such as image-based guidance or text prompts, to generative models for producing high-quality, user-specific 3D avatars. I am interested in learning controllable generative models with limited 3D supervision and incorporating large-scale generative priors to create diverse avatars.
Photo took in Kobe, Japan. 2016
Education
ETH Zürich
Advisor : Prof. Dr. Otmar Hilliges
[May.2022 - ]
National Taiwan University
[September.2012 - June.2016]
Tokyo Institute of Technology
Advisor : Prof. Hideki Koike
[September.2015 - March.2016]
[September.2009 - June.2012]
Practical Experience
Advanced Interactive Technologies Lab,
Department of Computer Science, ETH Zürich
Advisor : Dr. Jie Song, Prof. Dr. Otmar Hilliges
[June.2021 - February.2022]
Vision and Learning Lab, National Taiwan University /
Multimedia and Machine Learning Lab, Academia Sinica
Advisor : Prof. Yu-Chiang Frank Wang
[March.2017 - June.2018]
Media IC and System Lab, National Taiwan University
Advisor : Prof. Shao-Yi Chien
[September.2016 - Janruary.2017]
Publications
SiTH: Single-view Textured Human Reconstruction with Image-Conditioned Diffusion
[CVPR 2024 (BibTeX)] [Project Page] [Code] [Video]
4D-DRESS: A 4D Dataset of Real-world Human Clothing with Semantic Annotations
(* indicates equal contribution.)
[CVPR 2024 Highlight (BibTeX)] [Project Page] [Code] [Video]
Render In-between: Motion Guided Video Synthesis for Action Interpolation
[BMVC 2021 (BibTeX)] [Supplementary] [Code] [Video]
READ: Reciprocal Attention Discriminator for Image-to-Video Re-Identification
[ECCV 2020 (BibTeX)] [Video and Slides] [Code]
Learning from Dances: Pose-Invariant Re-Identification for Multi-Person Tracking
[ICASSP 2020 (BibTeX)] [Dataset] [Code] [Demo]
Summarizing First-Person Videos from Third Persons’ Points of View
[ECCV 2018 (BibTeX)] [Supplementary] [Project Page]
A Benchmark Dataset for 6DoF Object Pose Tracking
(* indicates equal contribution.)
[ISMAR 2017 (BibTeX)] [Project Page]
Selected Works
Course Projects
Learning Pose-aware Human Representations for Conditional Person Image Translation
Deep Learning course project at ETH Zürich
Spatiotemporal Multimodal Network for Dynamic Gesture Recognition
Machine Perception course project at ETH Zürich
Supe®oad: Road Segmentation through Multi-objective Ensemble and Geometric-aware Post-processing
(Names are in alphabetical order of first names.)
Computational Intelligence Lab course project at ETH Zürich
Neural Network Adversarial Attack Analysis
Reliable and Interpretable Artificial Intelligence course project at ETH Zürich
[Project Report] [Code]
Copyright © 2017 Husan-I Ho
Special Thanks : Walon Chiu provides his template
(Last Modified: 1st.5.2022)