In this work, we aim to obtain category-based 3D mesh models in .OBJ format through a Deep Neural Network from a single depth map. We introduce DepthOBJ, a synthetic dataset consisting of 3D mesh models divided in 54 categories and 19440 depth maps from 9 different angles in .PNG format. Recognizing the category in which the object depicted in the depth map belongs via a Convolutional Neural Network, we are able to produce the corresponding 3D mesh model in DepthOBJ using PyTorch3D.
Depthobj: A synthetic dataset for 3d mesh model retrieval
Snidaro L.
2021-01-01
Abstract
In this work, we aim to obtain category-based 3D mesh models in .OBJ format through a Deep Neural Network from a single depth map. We introduce DepthOBJ, a synthetic dataset consisting of 3D mesh models divided in 54 categories and 19440 depth maps from 9 different angles in .PNG format. Recognizing the category in which the object depicted in the depth map belongs via a Convolutional Neural Network, we are able to produce the corresponding 3D mesh model in DepthOBJ using PyTorch3D.File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.