I'm Jia Zheng, an engineer and researcher at Manycore Research, Manycore Tech Inc.
Projects
- SpatialLM: a large language model for spatial understanding.
- KoolCogVideoX: a text-to-video generative model for interior designs.
Selected Publications
- CAD2Program [AAAI 2025]: a vision-language model to 3D parametric model reconstruction from 2D CAD drawings.
- PlankAssembly [ICCV 2023]: the first deep generative models in the task of 3D CAD model reconstruction from three orthographic views.
- Faceformer [CVPR 2022]: the first data-driven approach to identify face loops from a single line drawing.
- NonCuboidRoom [WACV 2022]: a single-view 3D non-cuboid room layout reconstruction approach.
- Structured3D [ECCV 2020]: a large-scale photo-realistic dataset for data-driven structured 3D reconstruction.
- PlaneAE [CVPR 2019]: a single-view piece-wise 3D planar reconstruction approach.
Resources
- awesome-CAD: a list of awesome Computer-Aided Design (CAD) papers.
- awesome-scene-understanding: a list of scene understanding papers.
- awesome-holistic-3d: a list of holistic 3D reconstruction papers.
