img/jwgu5.jpg

Jinwei Gu

Principal Research Scientist

NVIDIA

[Google Scholar]
I am currently a principal research scientist at NVIDIA, working on deep generative models, vision foundation models, world models, and the general fields of computer vision, computer graphics, and machine learning. I am also an adjunct associate professor at the Chinese University of Hong Kong. I received my Ph.D. from Columbia University, and B.S. and M.S. from Tsinghua University.

Recent Research (to be updated ...)

adacode.png

Learning Image-Adaptive Codebooks for Class-Agnostic Image Restoration [Paper]

Kechun, Yitong, Inchang, Jinwei ICCV 2023


dtd.jpg

Real-time Controllable Denoising for Image and Video [Paper]

Zhaoyang, Yitong, Wenqi, Xiaogang, Ping, Kaimo, Jinwei. CVPR 2023 (to appear)


udc2.jpg

Generating Aligned Pseudo-Supervision from Non-Aligned Data for Image Restoration in Under-Display Camera[Paper]

Ruicheng, Chongyi, Huaijin, Shuai, Jinwei, and Cavan. CVPR 2023 (to appear)


fourier.jpg

Deep Fouier Up-Sampling[Paper]

Man, Yu, Jie, Feng, Jinwei, Cavan, Deyu, Chongyi. NeurIPS 2022


pinhole.jpg

Deep Camera Obscura: An Image Restoration Pipeline for Lenless Pinhole Photography[Paper]

Joshua, Huaijin, Shuai, Jinwei, and Suren. Optics Express, 2022


lowlight.jpg

Low-light Image and Video Enhancement Using Deep Learning: A Survey[Paper][Data+Code]

Chongyi, Chunle, Linghao, Jun, Ming-Ming, Jinwei and Chen Change. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021. (to appear)


itof2dtof.jpg

iToF2dToF: A Robust and Flexible Representation for Data-driven Time-of-Flight Imaging[Project(Paper+Data+Code)]

Felipe, Huaijin, Mohit, Andreas, and Jinwei. IEEE Transactions on Computational Imaging, 2021. (to appear)


reconfigISP.jpg

ReconfigISP: Reconfigurable Camera Image Processing Pipeline[Project (Paper+Code)]

Ke, Zexian, Yue, Cavan, and Jinwei. ICCV, 2021.


star.jpg

STAR: A Structure-Aware Lightweight Transformer for Real-time Image Enhancement [Paper]

Zhaoyang, Yitong, Jun, Xiaogang, Ping, and Jinwei ICCV, 2021.


quan.jpg

Dynamic Quantization with Mixed Precision and Adaptive Resolution [Paper]

Zhaoyang, Wenqi, Jinwei, Xiaogang, and Ping ICML, 2021. (spotlight)


udc.png

Removing Diffraction Image Artifacts in Under-Display Camera via Dynamic Skip Connection Network [Project/Code]

Ruicheng, Chongyi, Huaijin, Shuai, Cavan, and Jinwei CVPR, 2021.


glean.jpg

GLEAN: Generative Latent Bank for Large-Factor Image Super-Resolution [Project/Code]

Kelvin, Xintao, Xiangyu, Jinwei, and Cavan CVPR, 2021. (Oral)


vstitch.png

Video Stitching for Linear Camera Arrays [Project]

Wei-Sheng, Orazio, Jinwei, Deqing, Ming-hsuan, and Jan. BMVC, 2019. (to appear)


nrmvs.png

NRMVS: Non-Rigid Multi-View Stereo [Arxiv][PDF][Supplementary Video][Project (coming soon)]

Matthias, Kihwan, Jinwei, Matthias, Charles, Marc, and Jan. Arxiv, 2019.


rgbd.png

Neural RGB->D Sensing: Depth and Uncertainty from a Video Camera [Arxiv][Code]

Chao, Jinwei, Kihwan, Srinivas, and Jan. CVPR 2019 (Oral, Best Paper Finalist).


planercnn.png

PlaneRCNN: 3D Plane Detection and Reconstruction from a Single Image[Arxiv] [Code]

Chen, Kihwan, Jinwei, Yasutaka, and Jan. CVPR 2019 (Oral).


lightmat.png

Neural Inverse Rendering of an Indoor Scene from a Single Image[Arxiv][Project and Dataset]

Soumyadip, Jinwei, Kihwan, Guilin, David, and Jan. Arxiv 2019. ICCV 2019 (to appear)


tof.png

Very Power Efficient Neural Time-of-Flight[Arxiv][PDF][Supplementary Video][Project (coming soon)]

Yan, Jimmy, Xuanye, Keyuan, and Jinwei. Arxiv 2019.


insert.jpg

Context-Aware Synthesis and Placement of Object Instances[Arxiv][PDF][Supplementary Video][Code]

Donghoon, Sifei, Jinwei, Ming-Yu, Ming-Hsuan, and Jan. NIPS 2018.


tpn.png

Switchable Temporal Propagation Network [Arxiv][PDF][Supplementary Video][Applications]

Sifei, Jinwei, Shalini, Guangyu, Varun, Ming-Hsuan, and Jan. ECCV 2018.


reflect.png

Separating Reflection and Transmission Images in the Wild [Arxiv][PDF][Supplementary Video][Code]

Patrick, Orazio, Jinwei, and Jan. ECCV 2018.


mapnet.jpg

MapNet: Geometry-Aware Learning of Maps for Camera Localization[PDF][Code][Project]

Samarth, Jinwei, Kihwan, James, and Jan. CVPR 2018 (spotlight).


adapt_exposure.png

Space-Time-Brightness Sampling Using an Adaptive Pixel-Wise Coded Exposure [PDF]

Hajime, Toshiki, Dengyu, and Jinwei. CVPR 2018, Computational Camera and Display Workshop.


mvd.png

Reblur2Deblur: Deblurring Videos via Self-Supervised Learning[PDF][Project]

Huaijin, Jinwei, Orazio, Ming-Yu, Ashok, and Jan. ICCP 2018.


spn.jpg

Learning Affinity via Spatial Propagation Networks [PDF][Supplementary Video]

Sifei, Shalini, Jinwei, Guangyu, Ming-Hsuan,and Jan. NIPS 2017 (Accepted).


brdfnn.jpg

A Lightweight Approach for On-the-Fly Reflectance Estimation[PDF][Supplementary Video]

Kihwan, Jinwei, Stephen, Pavlo, Matthias, and Jan. ICCV 2017 Oral. arXiv:1705.07162


rnnface.jpg

Dynamic Facial Analysis: From Bayesian Filtering to Recurrent Neural Networks [Project][PDF][Supplementary Video]

Jinwei, Xiaodong, Shalini and Jan. CVPR 2017


pmvs.jpg

Polarimetric Multi-view Stereo [PDF][Supplementary Document][Video]

Zhaopeng, Jinwei, Boxin, Ping and Jan. CVPR 2017

evw17.jpg

Reconstructing Intensity Images from Binary Spatial Gradient Cameras [PDF][Supplementary Document][Video]

Suren, Orazio, Jinwei, Timo and Jan. CVPR EVW 2017 (Oral)

graincam.jpg

Deep Learning with Energy-efficient Binary Gradient Cameras [PDF]

Suren, Orazio, and Jinwei. arXiv:1612.00986, 2016

stcpw.jpg

Video Stitching with Spatial-Temporal Content-Preserving Warping[PDF][Video Results]

Wei and Jinwei. CVPR 2015 Workshop on Computational Cameras and Displays

top



Previous Research

fineart.jpg

Exemplar-Based Method for Automatic Visual Editing and Retouching of Fine Art Reproduction [project]

Jun and Jinwei. CIC 2013


texture_thumb.png

Learning Discriminative Illumination and Filters for BTF Classification[project] [database](with code)

Chao, Jinwei, and Gefei. CVPR 2013


camspec.jpg

What is the Space of Camera Spectral Sensitivities?[project] [database](with code)

Jun, Dengyu, Jinwei, and Sabine. IEEE WACV 2013


dome.jpg

Classifying Raw Materials with Discriminative Illumination [project] [database]

Chao and Jinwei. CVPR 2012 (Oral)


spectral.jpg

Spectral Imaging Under Commonly Available Lighting [project]

Jun and Jinwei. CVPR 2012 Workshop on Computational Camera and Displays


demux.jpg

Multiplexed Illumination for Scene Recovery in the Presence of Global Illumination[project]

Toshi, Jinwei, Mohit, and Shree. ICCV 2011 (Oral)


cpev.jpg

Video from a Single Exposure Coded Photograph using a Learned Over-Complete Dictionary[project]

Yas, Jinwei, Mohit, Tomoo, and Shree. ICCV 2011


crsp.jpg

Coded Rolling Shutter Photography: Flexible Space-Time Sampling [project]

Yas, Jinwei, Tomoo, and Shree. ICCP 2010


columbia.jpg

Measurement, Modeling, and Synthesis of Time-Varying Appearance of Natural Phenomena [PDF]

Jinwei Gu. PhD Thesis. Columbia University. March 2010


dirtylens.jpg

Removing Image Artifacts Due to Dirty Camera Lenses and Thin Occluders[project]

Jinwei, Ravi, Peter, and Shree. SIGGRAPH Asia 2009


ramp.jpg

Printing Spatially-Varying Reflectance [project]

Jinwei, Wojciech Matusik, Boris Ajdin, Jason Lawrence, Hendrik P. A. Lensch, Fabio Pellacini, and Szymon Rusinkiewicz. SIGGRAPH Asia 2009


csl.jpeg

Compressive Structured Light for Recovering Inhomogenous Participating Media [project](with code)

Jinwei, Ravi, Peter, Shree, and Eitan. ECCV 2008 (Oral)


dirtyglass.jpg

Dirty Glass: Modeling and Rendering Contamination on Transparent Surfaces [project] [database] (with code)

Jinwei, Ravi, Peter, and Shree. EGSR 2007


img/tvbtf.jpg

Time-Varying Surface Appearance [project][database] (with code)

Jinwei, Chien-i Tu, Ravi, Peter, Wojciech, and Shree. SIGGRAPH 2006

img/radiocalib.jpg

Radiometric Calibration from a Single Image [project]

Jinwei, Stephen Lin, Shuntaro Yamazaki, and Harry Shum, CVPR 2004 (Oral)


img/fingerprint_ori_model.jpg

Fingerprint Orientation Field Modeling and Its Applications [project]

with Jie Zhou, Chunyu Yang. CVPR 2003


top


Selected Publications (to be updated ...)

Computer Vision and Machine Learning

Computational Photography

Computer Graphics and Physics-based Computer Vision

Image Processing and Biometrics

Other Areas

top


Valid HTML 4.01 Transitional