Location Field Descriptors: Single Image 3D Model Retrieval in the Wild

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Abstract

We present Location Field Descriptors, a novel approach for single image 3D model retrieval in the wild. In contrast to previous methods that directly map 3D models and RGB images to an embedding space, we establish a common low-level representation in the form of location fields from which we compute pose invariant 3D shape descriptors. Location fields encode correspondences between 2D pixels and 3D surface coordinates and, thus, explicitly capture 3D shape and 3D pose information without appearance variations which are irrelevant for the task. This early fusion of 3D models and RGB images results in three main advantages: First, the bottleneck location field prediction acts as a regularizer during training. Second, major parts of the system benefit from training on a virtually infinite amount of synthetic data. Finally, the predicted location fields are visually interpretable and unblackbox the system. We …
Original languageEnglish
Title of host publication2019 International Conference on 3D Vision (3DV)
Pages583-593
Publication statusPublished - 2019

Fingerprint

pixel
prediction
method

Cite this

Grabner, A., Roth, P. M., & Lepetit, V. (2019). Location Field Descriptors: Single Image 3D Model Retrieval in the Wild. In 2019 International Conference on 3D Vision (3DV) (pp. 583-593)

Location Field Descriptors: Single Image 3D Model Retrieval in the Wild. / Grabner, Alexander; Roth, Peter M.; Lepetit, Vincent.

2019 International Conference on 3D Vision (3DV). 2019. p. 583-593.

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Grabner, A, Roth, PM & Lepetit, V 2019, Location Field Descriptors: Single Image 3D Model Retrieval in the Wild. in 2019 International Conference on 3D Vision (3DV). pp. 583-593.
Grabner A, Roth PM, Lepetit V. Location Field Descriptors: Single Image 3D Model Retrieval in the Wild. In 2019 International Conference on 3D Vision (3DV). 2019. p. 583-593
@inproceedings{c1efa02ae7044369a7c062551834c9fc,
title = "Location Field Descriptors: Single Image 3D Model Retrieval in the Wild",
abstract = "We present Location Field Descriptors, a novel approach for single image 3D model retrieval in the wild. In contrast to previous methods that directly map 3D models and RGB images to an embedding space, we establish a common low-level representation in the form of location fields from which we compute pose invariant 3D shape descriptors. Location fields encode correspondences between 2D pixels and 3D surface coordinates and, thus, explicitly capture 3D shape and 3D pose information without appearance variations which are irrelevant for the task. This early fusion of 3D models and RGB images results in three main advantages: First, the bottleneck location field prediction acts as a regularizer during training. Second, major parts of the system benefit from training on a virtually infinite amount of synthetic data. Finally, the predicted location fields are visually interpretable and unblackbox the system. We …",
author = "Alexander Grabner and Roth, {Peter M.} and Vincent Lepetit",
year = "2019",
language = "English",
pages = "583--593",
booktitle = "2019 International Conference on 3D Vision (3DV)",

}

TY - GEN

T1 - Location Field Descriptors: Single Image 3D Model Retrieval in the Wild

AU - Grabner, Alexander

AU - Roth, Peter M.

AU - Lepetit, Vincent

PY - 2019

Y1 - 2019

N2 - We present Location Field Descriptors, a novel approach for single image 3D model retrieval in the wild. In contrast to previous methods that directly map 3D models and RGB images to an embedding space, we establish a common low-level representation in the form of location fields from which we compute pose invariant 3D shape descriptors. Location fields encode correspondences between 2D pixels and 3D surface coordinates and, thus, explicitly capture 3D shape and 3D pose information without appearance variations which are irrelevant for the task. This early fusion of 3D models and RGB images results in three main advantages: First, the bottleneck location field prediction acts as a regularizer during training. Second, major parts of the system benefit from training on a virtually infinite amount of synthetic data. Finally, the predicted location fields are visually interpretable and unblackbox the system. We …

AB - We present Location Field Descriptors, a novel approach for single image 3D model retrieval in the wild. In contrast to previous methods that directly map 3D models and RGB images to an embedding space, we establish a common low-level representation in the form of location fields from which we compute pose invariant 3D shape descriptors. Location fields encode correspondences between 2D pixels and 3D surface coordinates and, thus, explicitly capture 3D shape and 3D pose information without appearance variations which are irrelevant for the task. This early fusion of 3D models and RGB images results in three main advantages: First, the bottleneck location field prediction acts as a regularizer during training. Second, major parts of the system benefit from training on a virtually infinite amount of synthetic data. Finally, the predicted location fields are visually interpretable and unblackbox the system. We …

M3 - Conference contribution

SP - 583

EP - 593

BT - 2019 International Conference on 3D Vision (3DV)

ER -