Home | Shenhan Qian

钱深瀚

Shenhan Qian

PhD student
Technical University of Munich

Bio

I am a PhD student in the Computer Vision Group led by Prof. Daniel Cremers at the Technical University of Munich. Previously, I spent a year in the Visual Computing Group under the supervision of Prof. Matthias Nießner. I received my Master’s degree in Computer Science from ShanghaiTech University, advised by Prof. Shenghua Gao, and my Bachelor’s degree in Automation from Hunan University.

My research focuses on the visual perception of the physical world, with an emphasis on understanding its geometry, appearance, motion, and articulation. I aim to develop computational models that enable machines to interpret and interact with complex real-world environments more effectively.

Publications

DynSUP: Dynamic Gaussian Splatting from An Unposed Image Pair

Weihang Li^*, Weirong Chen^*, Shenhan Qian, Jiajie Chen, Daniel Cremers, Haoang Li

arXiv 2024

project / arxiv / bibtex

Dynamic radiance field reconstruction from only two images, enabled by object-level bundle adjustment.

MeGA: Hybrid Mesh-Gaussian Head Avatar for High-Fidelity Rendering and Head Editing

Cong Wang, Di Kang, Heyi Sun, Shenhan Qian, Zixuan Wang, Linchao Bao, Songhai Zhang

CVPR 2025

project / arxiv / code / bibtex

Editable head avatar creation with triangle meshes for the face and 3DGS for hair.

GaussianAvatars: Photorealistic Head Avatars with Rigged 3D Gaussians

Shenhan Qian, Tobias Kirschstein, Liam Schoneveld, Davide Davoli^*, Simon Giebenhain, Matthias Nießner^†

CVPR 2024 (Highlight)

project / arxiv / video / code / face tracker / bibtex

Rigging 3D Gaussians to a parametric mesh model for photorealistic avatar creation and manipulation.

NeRSemble: Multi-view Radiance Field Reconstruction of Human Heads

Tobias Kirschstein, Shenhan Qian, Simon Giebenhain, Tim Walter, Matthias Nießner^†

SIGGRAPH 2023

project / arxiv / video / code / bibtex

Capturing the dynamics and appearance of human heads with a deformation field and an ensemble of radiance fields.

Cloth2Body: Generating 3D Human Body Mesh from 2D Clothing

Lu Dai^*, Liqian Ma^†, Shenhan Qian, Hao Liu, Ziwei Liu^†, Hui Xiong^†

ICCV 2023

project / paper / bibtex

Solving an inverse problem of virtual try-on with regression, retrieval, and diversification.

UNIF: United Neural Implicit Functions for Clothed Human Reconstruction and Animation

Shenhan Qian^*, Jiale Xu, Ziwei Liu, Liqian Ma^†, Shenghua Gao

ECCV 2022

project / arxiv / video / talk (Chinese) / code / poster / bibtex

Part-based human body reconstruction without part labels.

Dual-Space NeRF: Learning Animatable Avatars and Scene Lighting in Separate Spaces

Yihao Zhi^*, Shenhan Qian^*, Xinhao Yan^*, Shenghua Gao^†

3DV 2022

project / arxiv / video / code / bibtex

Modeling the human body and scene lighting in a pose-generalizable manner.

Feature Re-Representation and Reliable Pseudo Label Retraining for Cross-Domain Semantic Segmentation

Jing Li, Kang Zhou, Shenhan Qian, Wen Li, Lixin Duan, Shenghua Gao^†

T-PAMI 2022

paper / bibtex

Speech Drives Templates: Co-Speech Gesture Synthesis with Learned Templates

Shenhan Qian^*, Zhi Tu^*, Yihao Zhi^*, Wen Liu, Shenghua Gao^†

ICCV 2021

project / arxiv / paper / video / code / poster / bibtex

Generating upper-body gestures from voice with decoupled gesture styles and voice-compatible movements.

KGDet: Keypoint-Guided Fashion Detection

Shenhan Qian^*, Dongze Lian^*, Binqiang Zhao, Tong Liu, Bohui Zhu, Hai Li, Shenghua Gao^†

AAAI 2021

paper / code / bibtex

Improving clothing detection with positional and categoric clues of clothing landmarks.

Learning to Recommend Frame for Interactive Video Object Segmentation in the Wild

Zhaoyuan Yin, Jia Zheng, Weixin Luo, Shenhan Qian, Hanling Zhang^†, Shenghua Gao

CVPR 2021

arxiv / paper / code / bibtex

Predicting the most influential frame for additional annotation to boost interactive video object segmentation.

Projects

Versatile Head Alignment with Adaptive Appearance Priors

Sep, 2024

Open Source Project

code

A complete head tracking pipeline from videos to NeRF/3DGS-ready datasets.

DeepFashion2 Challenge (CVPR 2020 Workshop)

Apr, 2020

3rd Place in the Landmark Estimation Track

record / code

Combined a multi-stage detector (HTC) and a landmark estimator (HRNet) to detect fashion landmarks.

Lane Detection with An Instance Segmentation Approach

May, 2019

Bachelor's Thesis

code

A PyTorch implementation of LaneNet with instance segmentation and curve fitting.

College Student Electronics Design Contest of Hunan Province

Aug, 2018

1st Prize

Built a twin-rotor UAV that can take off, land, and hover steadily with 3d-printed parts, microcontrollers, and sensors. Achieved the highest precision of hovering height among all teams.

National College Student Electronics Design Contest

Aug, 2017

2nd Prize in Hunan Province

Built a tilting board system to control the movement of a ball on the board with microcontrollers, servos, and a camera. Accomplished smooth and flexible trajectory control.