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

Alexander Grabner, Peter M. Roth, Vincent Lepetit

Research output: Chapter in Book/Report/Conference proceedingConference paperpeer-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

Dive into the research topics of 'Location Field Descriptors: Single Image 3D Model Retrieval in the Wild'. Together they form a unique fingerprint.

Cite this