| CARVIEW |
Select Language
HTTP/1.1 200 OK
Date: Mon, 29 Dec 2025 01:12:46 GMT
Server: Apache/2.4.52 (Ubuntu)
Vary: Accept-Encoding
Content-Encoding: gzip
Cache-Control: max-age=0, no-cache, no-store, must-revalidate
Pragma: no-cache
Expires: Wed, 11 Jan 1984 05:00:00 GMT
Content-Length: 12113
Content-Type: text/html; charset=UTF-8
Xiaoming Zhao, PhD
Email / Google Scholar / GitHub / CV
I am a Research Scientist in the Machine Learning Research (MLR) team at Apple.
I earned my Ph.D. from Department of Computer Science, University of Illinois Urbana–Champaign (UIUC) under the guidance of Prof. Alexander Schwing.
Earlier, I received B.S. degree in Statistics from University of Science and Technology of China (USTC).
I am passionate about computer vision, generative models, and machine learning, with a broader goal of unifying understanding and generation within vision and beyond.
During my graduate study, I interned at Apple, Meta Reality Labs, and Google, conducting research related to above topics.
Email / Google Scholar / GitHub / CV
Publications
| [14] |
|
Studying Classifier(-Free) Guidance From a Classifier-Centric Perspective.
AAAI Conference on Artificial Intelligence (AAAI), 2026 Studying Classifier(-Free) Guidance From a Classifier-Centric Perspective
@inproceedings{zhao2025OnCFG,
|
| [13] |
|
3D Shape Tokenization via Latent Flow Matching.
arXiv, 2025 3D Shape Tokenization via Latent Flow Matching
@article{chang2025shapetoken,
|
| [12] |
|
IllumiNeRF: 3D Relighting Without Inverse Rendering.
Neural Information Processing Systems (NeurIPS), 2024
Media:
Radiance Fields
IllumiNeRF: 3D Relighting Without Inverse Rendering
@inproceedings{zhao2024illuminerf,
|
| [11] |
|
GoMAvatar: Efficient Animatable Human Modeling From Monocular Video Using Gaussians-on-Mesh.
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024 GoMAvatar: Efficient Animatable Human Modeling From Monocular Video Using Gaussians-on-Mesh
@inproceedings{wen2024GoM,
|
| [10] |
|
NeRFDeformer: NeRF Transformation From a Single View via 3D Scene Flows.
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024 NeRFDeformer: NeRF Transformation From a Single View via 3D Scene Flows
@inproceedings{tang2024nerfdeformer,
|
| [9] |
|
Pseudo-Generalized Dynamic View Synthesis From a Video.
International Conference on Learning Representations (ICLR), 2024 Pseudo-Generalized Dynamic View Synthesis From a Video
@inproceedings{Zhao2024PGDVS,
|
| [8] |
|
Occupancy Planes for Single-View RGB-D Human Reconstruction.
AAAI Conference on Artificial Intelligence (AAAI), 2023 Occupancy Planes for Single-View RGB-D Human Reconstruction
@inproceedings{Zhao2023Oplanes,
|
| [7] |
|
Generative Multiplane Images: Making a 2D GAN 3D-Aware.
European Conference on Computer Vision (ECCV), 2022 (Oral Presentation)
Media:
机器之心 (Chinese) /
MarkTechPost
Generative Multiplane Images: Making a 2D GAN 3D-Aware
@inproceedings{zhao2022gmpi,
|
| [6] |
|
Initialization and Alignment for Adversarial Texture Optimization.
European Conference on Computer Vision (ECCV), 2022 Initialization and Alignment for Adversarial Texture Optimization
@inproceedings{zhao2022tex,
|
| [5] |
|
Class-agnostic Reconstruction of Dynamic Objects From Videos.
(* denotes equal contribution) Neural Information Processing Systems (NeurIPS), 2021 Class-agnostic 4D Reconstruction From Videos
@inproceedings{ren2021redo,
|
| [4] |
|
The Surprising Effectiveness of Visual Odometry Techniques for Embodied PointGoal Navigation.
International Conference on Computer Vision (ICCV), 2021 The Surprising Effectiveness of Visual Odometry Techniques for Embodied PointGoal Navigation
@inproceedings{Zhao2021pointnav,
|
| [3] |
|
Mitigating Data Scarcity in Protein Binding Prediction Using Meta-Learning.
(* denotes equal contribution) Research in Computational Molecular Biology (RECOMB), 2019 Mitigating Data Scarcity in Protein Binding Prediction Using Meta-Learning
@inproceedings{luo2019mitigating,
|
| [2] |
|
Integrating Thermodynamic and Sequence Contexts Improves Protein-RNA Binding Prediction.
PLOS Computational Biology, 2019 Integrating Thermodynamic and Sequence Contexts Improves Protein-RNA Binding Prediction
@article{su2019integrating,
|
| [1] |
|
Stochastic Variance Reduction for Deep Q-Learning.
arXiv, 2019 Stochastic Variance Reduction for Deep Q-Learning
@article{Zhao2019RLarxiv,
|
Slides
|
Harnessing Data Priors to Mitigate 3D Data Scarcity.
The slides are almost the same as those for my job talk Harnessing "Dark" Data. 2024/10: PhD Thesis Defense |