Research  »

Deep Neural Models of Vegetation

We have developed algorithms that encode tree structure and behavior as a deep neural model. In particular, it encodes tree branchlets as a latent space that can be then used to create novel trees or reconstruct existing from point clouds.