We study 3D shape modeling from a single image and make contributions to it
in three aspects. First, we present Pix3D, a large-scale benchmark of diverse
image-shape pairs with pixel-level 2D-3D alignment. Pix3D has wide applications
in shape-related tasks including reconstruction, retrieval, viewpoint
estimation, etc. Building such a large-scale dataset, however, is highly
challenging; existing datasets either contain only synthetic data, or lack
precise alignment between 2D images and 3D shapes, or only have a small number
of images. Second, we calibrate the evaluation criteria for 3D shape
reconstruction through behavioral studies, and use them to objectively and
systematically benchmark cutting-edge reconstruction algorithms on Pix3D.
Third, we design a novel model that simultaneously performs 3D reconstruction
and pose estimation; our multi-task learning approach achieves state-of-the-art
performance on both tasks.
Описание
[1804.04610] Pix3D: Dataset and Methods for Single-Image 3D Shape Modeling
%0 Generic
%1 sun2018pix3d
%A Sun, Xingyuan
%A Wu, Jiajun
%A Zhang, Xiuming
%A Zhang, Zhoutong
%A Zhang, Chengkai
%A Xue, Tianfan
%A Tenenbaum, Joshua B.
%A Freeman, William T.
%D 2018
%K Pix3D
%T Pix3D: Dataset and Methods for Single-Image 3D Shape Modeling
%U http://arxiv.org/abs/1804.04610
%X We study 3D shape modeling from a single image and make contributions to it
in three aspects. First, we present Pix3D, a large-scale benchmark of diverse
image-shape pairs with pixel-level 2D-3D alignment. Pix3D has wide applications
in shape-related tasks including reconstruction, retrieval, viewpoint
estimation, etc. Building such a large-scale dataset, however, is highly
challenging; existing datasets either contain only synthetic data, or lack
precise alignment between 2D images and 3D shapes, or only have a small number
of images. Second, we calibrate the evaluation criteria for 3D shape
reconstruction through behavioral studies, and use them to objectively and
systematically benchmark cutting-edge reconstruction algorithms on Pix3D.
Third, we design a novel model that simultaneously performs 3D reconstruction
and pose estimation; our multi-task learning approach achieves state-of-the-art
performance on both tasks.
@misc{sun2018pix3d,
abstract = {We study 3D shape modeling from a single image and make contributions to it
in three aspects. First, we present Pix3D, a large-scale benchmark of diverse
image-shape pairs with pixel-level 2D-3D alignment. Pix3D has wide applications
in shape-related tasks including reconstruction, retrieval, viewpoint
estimation, etc. Building such a large-scale dataset, however, is highly
challenging; existing datasets either contain only synthetic data, or lack
precise alignment between 2D images and 3D shapes, or only have a small number
of images. Second, we calibrate the evaluation criteria for 3D shape
reconstruction through behavioral studies, and use them to objectively and
systematically benchmark cutting-edge reconstruction algorithms on Pix3D.
Third, we design a novel model that simultaneously performs 3D reconstruction
and pose estimation; our multi-task learning approach achieves state-of-the-art
performance on both tasks.},
added-at = {2019-02-09T20:04:28.000+0100},
author = {Sun, Xingyuan and Wu, Jiajun and Zhang, Xiuming and Zhang, Zhoutong and Zhang, Chengkai and Xue, Tianfan and Tenenbaum, Joshua B. and Freeman, William T.},
biburl = {https://www.bibsonomy.org/bibtex/2a2617927977b2d90296134cfc065ce01/mardukasoka},
description = {[1804.04610] Pix3D: Dataset and Methods for Single-Image 3D Shape Modeling},
interhash = {5b1ad27f11799de1f602537e422cc4fe},
intrahash = {a2617927977b2d90296134cfc065ce01},
keywords = {Pix3D},
note = {cite arxiv:1804.04610Comment: CVPR 2018. The first two authors contributed equally to this work. Project page: http://pix3d.csail.mit.edu},
timestamp = {2019-02-09T20:04:28.000+0100},
title = {Pix3D: Dataset and Methods for Single-Image 3D Shape Modeling},
url = {http://arxiv.org/abs/1804.04610},
year = 2018
}