PeopleofSuperintelligence

Researcher

Weiyue Wang

Weiyue Wang is a computer vision researcher whose work spans crowd-scene analysis, 3D reconstruction and autonomous-driving perception. At USC, she coauthored research on 3D shape inpainting, point-cloud segmentation and single-view reconstruction, then moved to Waymo Perception and worked on LiDAR and point-cloud detection. In 2025, she coauthored the Gemini 2.5 report and worked on a Gemini model that reached gold-medal problem-solving level at the International Mathematical Olympiad. Since 2025, she is an AI researcher at Meta.

Now at
Meta
Works
35
Citations
2,926
h-index
17
Badges

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. 2016

    Clustering motion in crowds

    In 2016, she coauthors a paper on detecting coherent motions in crowd scenes and using clustering to identify semantic regions.

  2. 2017

    Filling in 3D shapes

    In 2017, she coauthors a hybrid framework for 3D shape inpainting that combines a 3D generative adversarial network with a long-term recurrent convolutional network.

  3. 2018

    SGPN for point clouds

    At USC, she is first author of SGPN in 2018, a deep-learning framework for 3D object-instance segmentation on point clouds.

  4. 2019

    Single-view 3D reconstruction

    In 2019, she coauthors DISN with Qiangeng Xu and collaborators. The network generates detailed 3D meshes from single images by predicting signed distance fields.

  5. By 2021

    From USC to Waymo

    By 2021, she is a software engineer at Waymo Perception; earlier, she studies computer vision as a Ph.D. student in USC’s Computer Graphics and Immersive Technologies Lab under Ulrich Neumann.

  6. 2021

    RSN for LiDAR detection

    In 2021, she coauthors RSN, a real-time 3D object detector that predicts foreground points from range images and uses sparse convolutions to detect objects.

  7. 2022

    SWFormer for point clouds

    In 2022, she coauthors SWFormer, a sparse-window transformer for 3D object detection in point clouds. On the Waymo Open Dataset, it reaches 73.36 L2 mAPH for vehicles and pedestrians.

  8. Since 2025

    Gemini research and Meta

    In July 2025, Meta hires her from Google DeepMind, where she works on a Gemini model that demonstrates gold-medal-level problem solving at the International Mathematical Olympiad. She also coauthors the Gemini 2.5 report in 2025.

Career

  1. Since 2025MetaResearcher
  2. Until 2025Google DeepMindResearcher
  3. 2025–2026University of California, Santa Barbara
  4. 2021–2025Sun Yat-sen University
  5. WaymoSoftware Engineer

Education

  1. University of Southern California

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

  1. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes Weiyue Wang. Our estimates of where the 943 people shaping AI who worked at Google and Google DeepMind work now: 446 at Google and 452 across 278 other organizations.
  2. 16% of AI career moves in the 2020s went from a university to a companyIncludes Weiyue Wang. We estimate that 16% of AI career moves in the 2020s went from a university to a company; 29% of moves starting at a university went to a company, and 76% ended the university role.
  3. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Meta. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.
  4. Companies supply the most prior-work hires at 7 of the 9 frontier AI labsIncludes Meta. Our estimates of the organizations where 550 people worked directly before joining 9 frontier AI labs across 620 career transitions, comparing the main feeder organizations per lab.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Weiyue Wang yet. Send us a free-licensed photo.

  1. A diffusion and clustering-based approach for finding coherent motions and understanding crowd scenes arxiv.org. Checked 22 October 2026.
  2. Publications – CGIT Lab cgit.usc.edu. Checked 22 October 2026.
  3. Shape Inpainting using 3D Generative Adversarial Network and Recurrent Convolutional Networks arxiv.org. Checked 22 October 2026.
  4. SGPN: Similarity Group Proposal Network for 3D Point Cloud Instance Segmentation arxiv.org. Checked 22 October 2026.
  5. DISN: Deep Implicit Surface Network for High-quality Single-view 3D Reconstruction arxiv.org. Checked 22 October 2026.
  6. Tutorials - SMARTCOMP 2021 smartcomp2021.weebly.com. Checked 22 October 2026.
  7. RSN: Range Sparse Net for Efficient, Accurate LiDAR 3D Object Detection arxiv.org. Checked 22 October 2026.
  8. SWFormer: Sparse Window Transformer for 3D Object Detection in Point Clouds arxiv.org. Checked 22 October 2026.
  9. Meta, Microsoft hire top AI talent from DeepMind economictimes.indiatimes.com. Checked 22 October 2026.
  10. Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities arxiv.org. Checked 22 October 2026.