You Only Look at One: Category-Level Object Representations for Pose Estimation From a Single Example

Walter Goodwin, Ioannis Havoutis, Ingmar Posner
Proceedings of The 6th Conference on Robot Learning, PMLR 205:1435-1445, 2023.

Abstract

In order to meaningfully interact with the world, robot manipulators must be able to interpret objects they encounter. A critical aspect of this interpretation is pose estimation: inferring quantities that describe the position and orientation of an object in 3D space. Most existing approaches to pose estimation make limiting assumptions, often working only for specific, known object instances, or at best generalising to an object category using large pose-labelled datasets. In this work, we present a method for achieving category-level pose estimation by inspection of just a single object from a desired category. We show that we can subsequently perform accurate pose estimation for unseen objects from an inspected category, and considerably outperform prior work by exploiting multi-view correspondences. We demonstrate that our method runs in real-time, enabling a robot manipulator to rearrange previously unseen objects faithfully in terms of placement and orientation. Finally, we showcase our method in a continual learning setting, with a robot able to determine whether objects belong to known categories, and if not, use active perception to produce a one-shot category representation for subsequent pose estimation

Cite this Paper

BibTeX
@InProceedings{pmlr-v205-goodwin23a, title = {You Only Look at One: Category-Level Object Representations for Pose Estimation From a Single Example}, author = {Goodwin, Walter and Havoutis, Ioannis and Posner, Ingmar}, booktitle = {Proceedings of The 6th Conference on Robot Learning}, pages = {1435--1445}, year = {2023}, editor = {Liu, Karen and Kulic, Dana and Ichnowski, Jeff}, volume = {205}, series = {Proceedings of Machine Learning Research}, month = {14--18 Dec}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v205/goodwin23a/goodwin23a.pdf}, url = {https://proceedings.mlr.press/v205/goodwin23a.html}, abstract = {In order to meaningfully interact with the world, robot manipulators must be able to interpret objects they encounter. A critical aspect of this interpretation is pose estimation: inferring quantities that describe the position and orientation of an object in 3D space. Most existing approaches to pose estimation make limiting assumptions, often working only for specific, known object instances, or at best generalising to an object category using large pose-labelled datasets. In this work, we present a method for achieving category-level pose estimation by inspection of just a single object from a desired category. We show that we can subsequently perform accurate pose estimation for unseen objects from an inspected category, and considerably outperform prior work by exploiting multi-view correspondences. We demonstrate that our method runs in real-time, enabling a robot manipulator to rearrange previously unseen objects faithfully in terms of placement and orientation. Finally, we showcase our method in a continual learning setting, with a robot able to determine whether objects belong to known categories, and if not, use active perception to produce a one-shot category representation for subsequent pose estimation} }
Endnote
%0 Conference Paper %T You Only Look at One: Category-Level Object Representations for Pose Estimation From a Single Example %A Walter Goodwin %A Ioannis Havoutis %A Ingmar Posner %B Proceedings of The 6th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2023 %E Karen Liu %E Dana Kulic %E Jeff Ichnowski %F pmlr-v205-goodwin23a %I PMLR %P 1435--1445 %U https://proceedings.mlr.press/v205/goodwin23a.html %V 205 %X In order to meaningfully interact with the world, robot manipulators must be able to interpret objects they encounter. A critical aspect of this interpretation is pose estimation: inferring quantities that describe the position and orientation of an object in 3D space. Most existing approaches to pose estimation make limiting assumptions, often working only for specific, known object instances, or at best generalising to an object category using large pose-labelled datasets. In this work, we present a method for achieving category-level pose estimation by inspection of just a single object from a desired category. We show that we can subsequently perform accurate pose estimation for unseen objects from an inspected category, and considerably outperform prior work by exploiting multi-view correspondences. We demonstrate that our method runs in real-time, enabling a robot manipulator to rearrange previously unseen objects faithfully in terms of placement and orientation. Finally, we showcase our method in a continual learning setting, with a robot able to determine whether objects belong to known categories, and if not, use active perception to produce a one-shot category representation for subsequent pose estimation
APA
Goodwin, W., Havoutis, I. & Posner, I.. (2023). You Only Look at One: Category-Level Object Representations for Pose Estimation From a Single Example. Proceedings of The 6th Conference on Robot Learning, in Proceedings of Machine Learning Research 205:1435-1445 Available from https://proceedings.mlr.press/v205/goodwin23a.html.

Related Material