facebookresearch/meshrcnn
A deep learning model that reconstructs 3D meshes of objects from single 2D RGB images.
Not currently ranked — collecting fresh signals.
star history
Mesh R-CNN is a research implementation published at ICCV 2019 that predicts 3D polygonal meshes for objects detected in 2D images. It extends object detection baselines by adding a mesh prediction head that voxelizes or estimates vertex positions to reconstruct object geometry. The implementation builds on Detectron2 for object detection and PyTorch3D for 3D operations and rendering.
Frequently asked
- What is facebookresearch/meshrcnn?
- A deep learning model that reconstructs 3D meshes of objects from single 2D RGB images.
- Is meshrcnn open source?
- Yes — facebookresearch/meshrcnn is an open-source project tracked on heatdrop.
- What language is meshrcnn written in?
- facebookresearch/meshrcnn is primarily written in Python.
- How popular is meshrcnn?
- facebookresearch/meshrcnn has 1.2k stars on GitHub.
- Where can I find meshrcnn?
- facebookresearch/meshrcnn is on GitHub at https://github.com/facebookresearch/meshrcnn.