...

Gen Li

PhD student
CAB G82.1
gen.li@inf.ethz.ch
+41 44 632 06 38

Basic Information

I am a direct doctorate student at Computer Vision and Learning Group (VLG), ETH Zürich, supervised by Professor Siyu Tang. Before this, I received my Bachelor's degree in Computer Science at Tsinghua University (2020).

Social

Publications

EgoGen: An Egocentric Synthetic Data Generator

Conference: Conference on Computer Vision and Pattern Recognition (CVPR 2024) oral presentation

Authors:Gen Li, Kaifeng Zhao, Siwei Zhang, Xiaozhong Lyu, Mihai Dusmanu, Yan Zhang, Marc Pollefeys, Siyu Tang

EgoGen is new synthetic data generator that can produce accurate and rich ground-truth training data for egocentric perception tasks.
Project PDF Code BibTeX
TRTM: Template-based Reconstruction and Target-oriented Manipulation of Crumpled Cloths

Conference: IEEE International Conference on Robotics and Automation (ICRA 2024)

Authors:Wenbo Wang,  Gen Li, Miguel Zamora, and Stelian Coros

Precise reconstruction and manipulation of the crumpled cloths is challenging due to the high dimensionality of cloth models, as well as the limited observation at self-occluded regions. We leverage the recent progress in the field of single-view human reconstruction to template-based reconstruct crumpled cloths from their top-view depth observations only, with our proposed sim-real registration protocols. Our reconstruction mesh explicitly describes the positions and visibilities of the entire cloth mesh vertices, enabling more efficient dual-arm and single-arm target-oriented manipulations.
Project PDF Code BibTeX
Reconstructing Action-Conditioned Human-Object Interactions Using Commonsense Knowledge Priors

Conference: International Conference on 3D Vision (3DV 2022)

Authors:Xi Wang*,  Gen Li*, Yen-Ling Kuo, Muhammed Kocabas, Emre Aksan, and Otmar Hilliges
(* denotes equal contribution)

We present a method for inferring diverse 3D models of human-object interactions from images. We propose an action-conditioned modeling of interactions that allows us to infer diverse 3D arrangements of humans and objects without supervision on contact regions or 3D scene geometry. Our method extracts highlevel commonsense knowledge from large language models (such as GPT-3), and applies them to perform 3D reasoning of human-object interactions.
Project PDF BibTeX
Learning Topological Motion Primitives for Knot Planning

Conference: IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2020)

Authors:Mengyuan Yan,  Gen Li, Yilin Zhu, and Jeannette Bohg

We propose a hierarchical approach in which the top layer produces a topological plan and the bottom layer translates this plan into continuous robot motion.
Project PDF BibTeX