A Structure-Aware Method for Direct Pose Estimation

TL;DR

This paper introduces a structure-aware direct pose estimation method that incorporates explicit 3D constraints into a neural network, resulting in lower error compared to existing direct methods.

Contribution

It presents a novel direct pose estimation approach that integrates 3D structural information, improving accuracy over traditional direct methods like PoseNet.

Findings

Lower pose estimation error compared to PoseNet

Maintains constant-time inference

Incorporates explicit 3D constraints into the network

Abstract

Estimating camera pose from a single image is a fundamental problem in computer vision. Existing methods for solving this task fall into two distinct categories, which we refer to as direct and indirect. Direct methods, such as PoseNet, regress pose from the image as a fixed function, for example using a feed-forward convolutional network. Such methods are desirable because they are deterministic and run in constant time. Indirect methods for pose regression are often non-deterministic, with various external dependencies such as image retrieval and hypothesis sampling. We propose a direct method that takes inspiration from structure-based approaches to incorporate explicit 3D constraints into the network. Our approach maintains the desirable qualities of other direct methods while achieving much lower error in general.