Zhengqi Li (original) (raw)

I was a research scientist at Google DeepMind. I earned my Ph.D. in computer science from Cornell University, where I was advised by Professor Noah Snavely. I have received several awards for my research, including the 2020 Google Ph.D. Fellowship, the 2020 Adobe Research Fellowship, the 2021 Baidu Global Top 100 Rising Stars in AI, the CVPR 2019 and CVPR 2023 Best Paper Honorable Mention Award, the ICCV 2023 Best Student Paper Award, the CVPR 2024 Best Paper Award, CVPR 2025 Best Paper Award Candidate.

Resume/CV

News

Three paper was accepted to CVPR, 2025. MegaSaM was selected as Best Paper Award Candidate
I will serve as an Area Chair for CVPR 2025 and ICCV 2025.
One paper was accepted to SIGGRAPH, 2024.
Generative Image Dynamics was accepted to CVPR, 2024 and selected as Best Paper Award.
OmniMotion was accepted to ICCV, 2023 and selected as Best Student Paper Award .
Three paper was accepted to CVPR, 2023. DynIBaR was selected as Best Paper Honorable Mention
I will serve as an Area Chair for CVPR 2024.
Two paper was accepted to ECCV, 2022. One of them was selected as Oral Presentation
One paper was accepted to SIGGRAPH, 2022.
Three paper were accepted to CVPR, 2022. Two of them were selected as Oral Presentation
Our paper on dynamic scenes space-time view synthesis was accepted to CVPR, 2021.
I received 2020 Google PhD Fellowship. Thanks Google!
Our paper on Crowdsampling the Plenoptic Function was accepted to ECCV, 2020 and was selected as Oral Presentation.
Our extended paper on learning depth of moving people was accepted to TPAMI.
I received 2020 Adobe Research Fellowship (10 selected worldwide). Thanks Adobe!
Our paper on geometry aware camera orientation estimation was accepted to ICCV, 2019.
Our paper on learning the depth of dynamic scene with moving people was selected as CVPR 2019 Best Paper Honorable Mention
Our paper on learning intrinsic image decompostion from CGIntrinsics was accepted to ECCV, 2018.
Our paper on unsupervised learning intrinsic image decomposition was accepted to CVPR, 2018 and was selcted for a spolight oral.
Our paper on learning single-view depth prediction from Internet photos was accepted to CVPR, 2018

Publications/Technical Reports

Your browser does not support the video tag.

MegaSaM: Accurate, Fast and Robust Structure and Motion from Casual Dynamic Videos

Zhengqi Li, Richard Tucker, Forrester Cole, Qianqian Wang, Linyi Jin, Vickie Ye, Angjoo Kanazawa, Aleksander Holynski, Noah Snavely
In CVPR, 2025 (Best Paper Award Candidate)
[www]

Your browser does not support the video tag.

MegaDepth: Learning Single-View Depth Prediction from Internet Photos

Zhengqi Li and Noah Snavely

In CVPR, 2018 (also invited to be presented at Bridges to 3D Workshop, CVPR 2018)

[paper] [www] [Demo]

Media Coverage

Two Minute Papers: Google’s New AI: Fly INTO Photos… But Deeper!
Two Minute Papers: This AI Learned To Stop Time!
Cornell Chronicle Tool transforms world landmark photos into 4D experiences
Google AI Blog: Moving Camera, Moving People: A Deep Learning Approach to Depth Prediction
MIT Technology Review: If you did the Mannequin Challenge, you are now advancing robotics research
Two Minute Papers: This AI Learns About Movement By Watching Frozen People
VentureBeat: Google Leverages on YouTube Videos of Mannequin Challenge to Improve Depth Prediction
Seamless: Google、シーン内の人とカメラの両方が移動していても1台の単眼カメラから深度を予測するdeep learningを用いた手法を発表
量子位: 谷歌预测景深新研究：即使相机人物都在动，单一视点也能合成3D深度图
Robinly: CVPR2019爆款论文作者现场解读：视觉语言导航、运动视频深度预测、6D姿态估计
雷锋网: 单个运动摄像头估计运动物体深度，谷歌挑战新难题
雷锋网: 通过网络无标注延时摄影学习本征图像分解