Mesh R-CNN
@inproceedings{gkioxari2019mesh,
title={Mesh r-cnn},
author={Gkioxari, Georgia and Malik, Jitendra and Johnson, Justin},
booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
pages={9785--9795},
year={2019}
}
- Predict coarse shape occupancy Voxel Grid for every detected image in the scene
- Convert coarse shape into a Meshes so it can be used as a template like in Pixel2Mesh
- Rest of the procedure similar as Pixel2Mesh