Saved in:
Bibliographic Details
Main Authors: Sapoutzoglou, Panagiotis, Giapitzakis, Georgios, Floros, Georgios, Terzakis, George, Pateraki, Maria
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2404.16471
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929631463997440
author Sapoutzoglou, Panagiotis
Giapitzakis, Georgios
Floros, Georgios
Terzakis, George
Pateraki, Maria
author_facet Sapoutzoglou, Panagiotis
Giapitzakis, Georgios
Floros, Georgios
Terzakis, George
Pateraki, Maria
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2404_16471
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle COBRA -- COnfidence score Based on shape Regression Analysis for method-independent quality assessment of object pose estimation from single images
Sapoutzoglou, Panagiotis
Giapitzakis, Georgios
Floros, Georgios
Terzakis, George
Pateraki, Maria
Computer Vision and Pattern Recognition
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.
title COBRA -- COnfidence score Based on shape Regression Analysis for method-independent quality assessment of object pose estimation from single images
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2404.16471