COBRA -- COnfidence score Based on shape Regression Analysis for method-independent quality assessment of object pose estimation from single images

TL;DR

This paper introduces COBRA, a shape-based confidence scoring method for evaluating 6D object pose estimates from single images, using Gaussian Process shape templates to assess geometric discrepancies.

Contribution

It presents a novel, method-independent confidence measure for pose estimation based on shape discrepancy evaluation with Gaussian Process templates.

Findings

Shape templates accurately represent object geometry.

Confidence scores correlate with pose estimation quality.

Method is robust across different object types.

Abstract

We propose a generic procedure for assessing 6D object pose estimates. Our approach relies on the evaluation of discrepancies in the geometry of the observed object, in particular its respective estimated back-projection in 3D, against a putative functional shape representation comprising mixtures of Gaussian Processes, that act as a template. Each Gaussian Process is trained to yield a fragment of the object's surface in a radial fashion with respect to designated reference points. We further define a pose confidence measure as the average probability of pixel back-projections in the Gaussian mixture. The goal of our experiments is two-fold. a) We demonstrate that our functional representation is sufficiently accurate as a shape template on which the probability of back-projected object points can be evaluated, and, b) we show that the resulting confidence scores based on these probabilities are indeed a consistent quality measure of pose.