👩‍🔬 Renie Wei (魏榕)

Welcome! I am currently working as a Research Algorithm Expert at Manycore Tech Inc., focusing on 3D vision, robotics simulation, and generating synthetic data.

📧 wei_rong_pku[at]163.com
📍 Based in Hangzhou, China

---

🗞 News

• [2025.07] We release InteriorGS, a high-quality 3D Gaussian Splatting dataset of 1,000 richly annotated indoor scenes!
  InteriorGS bridges the gap between scan-based and artist-designed datasets by combining photorealistic rendering with 3DGS reconstruction from over 5 million images. It includes:
  • 554,000+ object instances across 755 categories
  • Per-object semantic labels and 3D bounding boxes
  • Occupancy maps for agent navigation
  • 80+ environment types (e.g., homes, museums, convenience stores)

  The dataset supports research in 3D scene understanding, controllable generation, and embodied AI. A web-based viewer is also available for exploring the data.

  Sample from the InteriorGS dataset.

💼 Work Experience

Manycore Tech Inc

Research Algorithm Expert | Present

Shuyuan Digital City Digital Naive Research Center

Senior Engineer | Jul 2023 – Mar 2025

Huawei Central Media Research Institute

Postdoctoral Researcher | Aug 2021 – Jul 2023

🎓 Education

Peking University
Ph.D. in Biomedical Engineering (3D data denoising and compressing)
2016.07 – 2021.07, Beijing, China

Xidian University
B.Eng. in Integrated Circuit Design and Systems
2012.07 – 2016.07, Xi’an, China

🧪 Research Projects (Selected)

• Layout Generation Framework: A framework for multimodal-conditioned 3D indoor scene generation using hierarchical scene graphs for controllable layout and semantics. Ongoing project, 2025
• SomaSeg Framework: Robust neuron segmentation for two-photon imaging videos. Journal of Neural Engineering, 2024
• PSC Differentiation Monitoring: Early-stage stem cell classification via brightfield imaging. Cell Discovery, 2023
• Explainable Prostate Cancer Detection: Multi-modal segmentation and registration for MRI. ISMRM, JMRI, etc.

---

📌 Patents (Partial List)

• CN-118154713-A: Scene rendering method and system
• CN-118052926-B: Image rendering and compression system
• WO-2024152959-A1: Cloud-edge collaborative 3D rendering encoding
• CN-114331900-A: Video denoising and reconstruction using low-rank modeling

---

🏆 Awards

• Outstanding Postdoctoral Researcher, Huawei Central Media Research Institute, 2022
• Gold Prize, China “Internet+” Innovation and Entrepreneurship Competition, 2018
• Meritorious Winner, MCM/ICM Mathematical Contest in Modeling, 2015

---

🌐 Skills & Languages

• Programming: C++, Python, PyTorch, Vulkan, UE5, Isaac Sim
• Languages: Chinese (native), English (CET-6)