Learning hierarchical relationships for object-goal navigation

Anwesan Pal, Yiding Qiu, Henrik Christensen
Proceedings of the 2020 Conference on Robot Learning, PMLR 155:517-528, 2021.

Abstract

Direct search for objects as part of navigation poses a challenge for small items. Utilizing context in the form of object-object relationships enable hierarchical search for targets efficiently. Most of the current approaches tend to directly incorporate sensory input into a reward-based learning approach, without learning about object relationships in the natural environment, and thus generalize poorly across domains. We present Memory-utilized Joint hierarchical Object Learning for Navigation in Indoor Rooms (MJOLNIR), a target-driven navigation algorithm, which considers the inherent relationship between target objects, and the more salient contextual objects occurring in its surrounding. Extensive experiments conducted across multiple environment settings show an 82.9% and 93.5% gain over existing state-of-the-art navigation methods in terms of the success rate (SR), and success weighted by path length (SPL), respectively. We also show that our model learns to converge much faster than other algorithms, without suffering from the well-known overfitting problem. Additional details regarding the supplementary material and code are available at https://sites.google.com/eng.ucsd.edu/mjolnir.

Cite this Paper

BibTeX
@InProceedings{pmlr-v155-pal21a, title = {Learning hierarchical relationships for object-goal navigation}, author = {Pal, Anwesan and Qiu, Yiding and Christensen, Henrik}, booktitle = {Proceedings of the 2020 Conference on Robot Learning}, pages = {517--528}, year = {2021}, editor = {Kober, Jens and Ramos, Fabio and Tomlin, Claire}, volume = {155}, series = {Proceedings of Machine Learning Research}, month = {16--18 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v155/pal21a/pal21a.pdf}, url = {https://proceedings.mlr.press/v155/pal21a.html}, abstract = {Direct search for objects as part of navigation poses a challenge for small items. Utilizing context in the form of object-object relationships enable hierarchical search for targets efficiently. Most of the current approaches tend to directly incorporate sensory input into a reward-based learning approach, without learning about object relationships in the natural environment, and thus generalize poorly across domains. We present Memory-utilized Joint hierarchical Object Learning for Navigation in Indoor Rooms (MJOLNIR), a target-driven navigation algorithm, which considers the inherent relationship between target objects, and the more salient contextual objects occurring in its surrounding. Extensive experiments conducted across multiple environment settings show an 82.9% and 93.5% gain over existing state-of-the-art navigation methods in terms of the success rate (SR), and success weighted by path length (SPL), respectively. We also show that our model learns to converge much faster than other algorithms, without suffering from the well-known overfitting problem. Additional details regarding the supplementary material and code are available at https://sites.google.com/eng.ucsd.edu/mjolnir.} }
Endnote
%0 Conference Paper %T Learning hierarchical relationships for object-goal navigation %A Anwesan Pal %A Yiding Qiu %A Henrik Christensen %B Proceedings of the 2020 Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2021 %E Jens Kober %E Fabio Ramos %E Claire Tomlin %F pmlr-v155-pal21a %I PMLR %P 517--528 %U https://proceedings.mlr.press/v155/pal21a.html %V 155 %X Direct search for objects as part of navigation poses a challenge for small items. Utilizing context in the form of object-object relationships enable hierarchical search for targets efficiently. Most of the current approaches tend to directly incorporate sensory input into a reward-based learning approach, without learning about object relationships in the natural environment, and thus generalize poorly across domains. We present Memory-utilized Joint hierarchical Object Learning for Navigation in Indoor Rooms (MJOLNIR), a target-driven navigation algorithm, which considers the inherent relationship between target objects, and the more salient contextual objects occurring in its surrounding. Extensive experiments conducted across multiple environment settings show an 82.9% and 93.5% gain over existing state-of-the-art navigation methods in terms of the success rate (SR), and success weighted by path length (SPL), respectively. We also show that our model learns to converge much faster than other algorithms, without suffering from the well-known overfitting problem. Additional details regarding the supplementary material and code are available at https://sites.google.com/eng.ucsd.edu/mjolnir.
APA
Pal, A., Qiu, Y. & Christensen, H.. (2021). Learning hierarchical relationships for object-goal navigation. Proceedings of the 2020 Conference on Robot Learning, in Proceedings of Machine Learning Research 155:517-528 Available from https://proceedings.mlr.press/v155/pal21a.html.

Related Material