Yinda Zhang

I am a Research Scientist at Google. My research interests lie at the intersection of computer vision, computer graphics, and machine learning. Recently, I focus on empowering 3D vision and perception via machine learning, including dense depth estimation, 3D shape analysis, and 3D scene understanding. I received my Ph.D. in Computer Science from Princeton University, advised by Professor Thomas Funkhouser. Before that, I received a Bachelor degree from Dept. Automation in Tsinghua University, and a Master degree from Dept. ECE in National University of Singapore co-supervised by Prof. Ping Tan and Prof. Shuicheng Yan.

Email:	yindanospamz (at) gmail (dot) com

Find me on:	Google Scholar GitHub LinkedIn

YES! YES! YES! YES! YES↑ YES! YES! YES!

2021.03	Five papers (2 Orals + 3 Posters) are accepted by CVPR2021.
2020.07	Three papers (2 Orals + 1 Posters) are accepted by ECCV2020.
2020.04	Our deep learning depth refinement solution is checked in Pixel4 front-facing camera to support computational photography and AR applications. Learn more here: Google AI blogpost.
2020.02	Pixel2Mesh is accepted by IEEE Transactions of Pattern Analysis and Machine Intelligence.
2020.02	Four papers are accepted by CVPR2020.
2019.12	Our deep learning based solution for image based depth estimation has been deloyed on Google Pixel4 for Portrait Mode. Check this Google AI blogpost for more details.
2019.07	Pixel2Mesh++ is accepted by ICCV2019. Check here for the paper.
2019.03	DeepLidar paper is accepted by CVPR2019.
2018.12	I started working at Google as a Research Scientist.
2018.11	I obtain Ph.D degree from Princeton University.
2018.11	I am awarded as Siebel Scholar Class of 2019.
2018.07	ActiveStereoNet is accepted as oral presentation by ECCV 2018.
2018.07	Pixel2Mesh is accepted by ECCV 2018.

HumanGPS: Geodesic PreServing Feature for Dense Human Correspondences

F. Tan, D. Tang, M. Dou, K. Guo, R. Pandey, C. Keskin, R. Du, D. Sun, S. Bouaziz, S. Fanello, P. Tan, Y. Zhang

Computer Vision and Pattern Recognition (CVPR 2021)

[Paper] [Project Webpage] [Codes]

Holistic 3D Scene Understanding from a Single Image with Implicit Representation

C. Zhang^*, Z. Cui^*, Y. Zhang^*, B. Zeng, M. Pollefeys, S. Liu

Computer Vision and Pattern Recognition (CVPR 2021)

[Paper] [Project Webpage] [Codes]

Learning Compositional Representation for 4D Captures with Neural ODE

B. Jiang^*, Y. Zhang^*, X. Wei, X. Xue, Y. Fu.

Computer Vision and Pattern Recognition (CVPR 2021)

[Paper] [Project Webpage] [Data, Model, Code] [Video]

Spatially-Varying Outdoor Lighting Estimation from Intrinsics

Y. Zhu, Y. Zhang, S. Li, B. Shi

Computer Vision and Pattern Recognition (CVPR 2021), Oral Presentation

[Paper]

HITNet: Hierarchical Iterative Tile Refinement Network for Real-time Stereo Matching

V. Tankovich, C. Hane, Y. Zhang, A. Kowdle, S. Fanello, S. Bouaziz

Computer Vision and Pattern Recognition (CVPR 2021), Oral Presentation

[Paper] [Pretrained Model and Evaluation Code]

Du2Net: Learning Depth Estimation from Dual-Cameras and Dual-Pixels

Y. Zhang, N. Wadhwa, S. Orts-Escolano, C. haene, S. Fanello, R. Garg

European Conference on Computer Vision (ECCV 2020), Oral Presentation

[Paper] [Project Webpage]

DeepSFM: Structure From Motion Via Deep Bundle Adjustment

X. Wei^*, Y. Zhang^*, Z. Li^*, Y. Fu, X. Xue

European Conference on Computer Vision (ECCV 2020), Oral Presentation

[Paper] [Project Webpage] [Codes]

GeoLayout: Geometry Driven Room Layout Estimation Based on Depth Maps of Planes

W. Zhang, W. Zhang, Y. Zhang

European Conference on Computer Vision (ECCV 2020)

[Paper] [Project Webpage] [Matterport3D-Layout Dataset]

PBR-Net: Imitating Physically Based Rendering using Deep Neural Network

P. Dai, Z. Li, Y. Zhang, S. Liu, B. Zeng

IEEE Transactions on Image Processing, 16 Apr 2020, DOI: 10.1109/TIP.2020.2987169.

[Paper]

Neural Pose Transfer by Spatially Adaptive Instance Normalization

J. Wang, C. Wen, Y. Fu, H. Lin, T. Zou, X. Xue, Y. Zhang

Computer Vision and Pattern Recognition (CVPR 2020)

[Paper] [Project Webpage (Poster, Video)] [Code and Data]

Deep Implicit Volume Compression

D. Tang, S. Singh, P. Chou, C. Haene, M. Dou, S. Fanello, J. Taylor, P. Davidson, O. Guleryuz, Y. Zhang, S. Izadi, A. Tagliasacchi, S. Bouaziz, C. Keskin

Computer Vision and Pattern Recognition (CVPR 2020), Oral Presentation

[Paper] [Project Webpage] [Video]

Neural Point Cloud Rendering via Multi-Plane Projection

P. Dai^*, Y. Zhang^*, Z. Li^*, S. Liu, B. Zeng

Computer Vision and Pattern Recognition (CVPR 2020)

[Paper] [Project Webpage] [Code and Data] [Poster] [Video]

DIST: Rendering Deep Implicit Signed Distance Function with Differentiable Sphere Tracing

S. Liu, Y. Zhang, S. Peng, B. Shi, M. Pollefeys, Z. Cui

Computer Vision and Pattern Recognition (CVPR 2020)

[Paper] [Supplementary Material] [Code and Data] [Project Webpage] [Video] [Poster] [Slides]

Pixel2Mesh++: Multi-View 3D Mesh Generation via Deformation

C. Wen^*, Y. Zhang^*, Z. Li^*, Y. Fu

International Conference on Computer Vision (ICCV 2019)

[Paper] [Code and Data] [Project Webpage]

DeepLiDAR: Deep Surface Normal Guided Depth Prediction for Outdoor Scene from Sparse LiDAR Data and Single Color Image

J. Qiu^*, Z. Cui^*, Y. Zhang^*, X. Zhang, S. Liu, B. Zeng, M. Pollefeys

Computer Vision and Pattern Recognition (CVPR 2019)

[Paper] [Code and Data]

ActiveStereoNet: End-to-End Self-Supervised Learning for Active Stereo Systems

Y. Zhang, S. Khamis, C. Rhemann, J. Valentin, A. Kowdle, V. Tankovich, M. Schoenberg, S. Izadi, T. Funkhouser, S. Fanello

European Conference on Computer Vision (ECCV 2018), Oral Presentation

[Paper] [Supplimentary] [Arxiv] [Slides] [Oral Presentation] [Project Webpage]

Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images

N. Wang^*, Y. Zhang^*, Z. Li^*, Y. Fu, W. Liu, Y. Jiang

European Conference on Computer Vision (ECCV 2018)

IEEE Transactions on Pattern Analysis and Machine Intelligence, 02 Apr 2020, DOI: 10.1109/TPAMI.2020.2984232.

[Paper] [Arxiv] [Code] [Project Webpage]

Deep Depth Completion of a Single RGB-D Image

Y. Zhang, T. Funkhouser

Computer Vision and Pattern Recognition (CVPR 2018)

[Paper] [Supplimentary] [Arxiv] [Spotlight Presentation] [Project Webpage] [Code]

Matterport3D: Learning from RGB-D Data in Indoor Environments

A. Chang, A. Dai, T. Funkhouser, M. Halber, M. Niessner, M. Savva, S. Song, A. Zeng, Y. Zhang

International Conference on 3D Vision (3DV 2017)

[Paper] [Arxiv] [Project Webpage]

Physically-Based Rendering for Indoor Scene Understanding Using Convolutional Neural Networks

Y. Zhang, S. Song, E. Yumer, M. Savva, J. Lee, H. Jin, T. Funkhouser

Computer Vision and Pattern Recognition (CVPR 2017)

[Paper] [Supplimentary] [Arxiv] [Project Webpage] [Code]

DeepContext: Context-Encoding Neural Pathways for 3D Holistic Scene Understanding

Y. Zhang, M. Bai, P. Kohli, S. Izadi, J. Xiao

International Conference on Computer Vision (ICCV 2017)

[Paper] [Supplimentary] [Arxiv] [Project Webpage] [Video Youtube] [Video Download]

Hand3D: Hand Pose Estimation using 3D Neural Network

X. Deng^*, S. Yang^*, Y. Zhang^*, P. Tan, L. Chang, H. Wang

arXiv:1704.02224 [cs.CV] (7 Apr 2017)

[Paper] [Project Webpage]

Joint Hand Detection and Rotation Estimation Using CNN

X. Deng, Y. Zhang, S. Yang, P. Tan, L. Chang, Y. Yuan, H. Wang

IEEE Transactions on Image Processing, 04 Dec 2017, DOI: 10.1109/TIP.2017.2779600.

[Paper] [Project Webpage]

LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop

F. Yu, A. Seff, Y. Zhang, S. Song, T. Funkhouser, J. Xiao

arXiv:1506.03365 [cs.CV] 10 Jun 2015

[Paper] [Project Webpage]

TurkerGaze: Crowdsourcing Saliency with Webcam based Eye Tracking

P. Xu, K. A. Ehinger, Y. Zhang, A. Finkelstein, S. R. Kulkarni, J. Xiao

arXiv:1504.06755 [cs.CV] 25 Apr 2015

[Paper] [Supplimentary] [Video Youtube] [Video Download] [Project Webpage]

PanoContext: A Whole-room 3D Context Model for Panoramic Scene Understanding

Y. Zhang, S. Song, P. Tan, and J. Xiao

European Conference on Computer Vision (ECCV 2014), Oral Presentation

[Paper] [Supplimentary] [Oral Presentation] [Conference Talk] [Detailed Video Youtube] [Detailed Video Download (8min)] [Short Video Download (1min)] [Project Webpage]

FrameBreak: Dramatic Image Extrapolation by Guided Shift-Maps

Y. Zhang, J. Xiao, J. Hays, P. Tan

Computer Vision and Pattern Recognition (CVPR 2013)

[Paper] [Supplimentary] [Project Webpage]

Google, Mountain View

Research Scientist
12/2018--Present

Princeton University, Vision and Robotics Group

Ph.D. in Department of Computer Science
09/2014--11/2018

Google (through AutoRoboto Inc.), Mountain View

Research Intern
09/2017--10/2018

Matterport Inc., Sunnyvale

Research Intern
06/2017--08/2017

Adobe Research, San Jose

Research Intern
06/2016--08/2016

Microsoft Research, Redmond

Research Intern
07/2015--08/2015

National University of Singapore, Singapore

M.Eng. in Department of Electrical and Computer Engineering
01/2010--01/2013

Microsoft Research Asia, Beijing, Vision Computing Group

Research Intern
05/2010--08/2011

Tsinghua University, Beijing

B.Eng. in Department of Automation
08/2005--08/2009

Invited Talk:

Organizer of Workshops & Tutorial:

Large-scale Scene Understanding Challenge Workshop (CVPR 2016)
Large-scale Scene Understanding Challenge Workshop (CVPR 2015)
Tutorial on 3D Deep Learning with Marvin (CVPR 2016)

Reviewer of Conferences:

CVPR, ECCV, ICCV, SIGGRAPH, NIPS, AAAI, BMVC, 3DV, ACCV, ICPR, ICRA

Reviewer of Journals:

PAMI, IJCV, TOG, MVAP, TIP, NEUCOM

Awards:

Siebel Scholars, Class of 2019
Outstanding Reviewer in CVPR2018
Silver Medal for National Olympic Competition of Physics, China, 2004

Teaching:

Preceptor and Teaching Assistant: Princeton COS435 Information Retrival, 2016 Spring
Preceptor and Teaching Assistant: Princeton COS429 Computer Vision, 2015 Fall

Previously Monitored Students:

Shaohui Liu, University of Washington/Tsinghua University.
Jiashun Wang, University of California San Diego/Fudan University.
Chao Wen, Fudan University.
Nanyang Wang, Alibaba.
Peng Dai, University of Hong Kong.
Jiaxiong Qiu, University of Electronic Science and Technology of China.
Ye Yuan, Face++.

Email:	yindanospamz (at) gmail (dot) com

Address:	1600 Amphitheatre Pkwy,
	Mountain View, CA, US, 94043

Yinda Zhang

News

Publications

HumanGPS: Geodesic PreServing Feature for Dense Human Correspondences

Holistic 3D Scene Understanding from a Single Image with Implicit Representation

Learning Compositional Representation for 4D Captures with Neural ODE

Spatially-Varying Outdoor Lighting Estimation from Intrinsics

HITNet: Hierarchical Iterative Tile Refinement Network for Real-time Stereo Matching

Du2Net: Learning Depth Estimation from Dual-Cameras and Dual-Pixels

DeepSFM: Structure From Motion Via Deep Bundle Adjustment

GeoLayout: Geometry Driven Room Layout Estimation Based on Depth Maps of Planes

PBR-Net: Imitating Physically Based Rendering using Deep Neural Network

Neural Pose Transfer by Spatially Adaptive Instance Normalization

Deep Implicit Volume Compression

Neural Point Cloud Rendering via Multi-Plane Projection

DIST: Rendering Deep Implicit Signed Distance Function with Differentiable Sphere Tracing

Pixel2Mesh++: Multi-View 3D Mesh Generation via Deformation

DeepLiDAR: Deep Surface Normal Guided Depth Prediction for Outdoor Scene from Sparse LiDAR Data and Single Color Image

ActiveStereoNet: End-to-End Self-Supervised Learning for Active Stereo Systems

Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images

Deep Depth Completion of a Single RGB-D Image

Matterport3D: Learning from RGB-D Data in Indoor Environments

Physically-Based Rendering for Indoor Scene Understanding Using Convolutional Neural Networks

DeepContext: Context-Encoding Neural Pathways for 3D Holistic Scene Understanding

Hand3D: Hand Pose Estimation using 3D Neural Network

Joint Hand Detection and Rotation Estimation Using CNN

LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop

TurkerGaze: Crowdsourcing Saliency with Webcam based Eye Tracking

PanoContext: A Whole-room 3D Context Model for Panoramic Scene Understanding

FrameBreak: Dramatic Image Extrapolation by Guided Shift-Maps

Experiences

Google, Mountain View

Princeton University, Vision and Robotics Group

Google (through AutoRoboto Inc.), Mountain View

Matterport Inc., Sunnyvale

Adobe Research, San Jose

Microsoft Research, Redmond

National University of Singapore, Singapore

Microsoft Research Asia, Beijing, Vision Computing Group

Tsinghua University, Beijing

Services and Awards

Invited Talk:

Organizer of Workshops & Tutorial:

Reviewer of Conferences:

Reviewer of Journals:

Awards:

Teaching:

Previously Monitored Students:

Contacts