Monocular Teach-and-Repeat Navigation using a Deep Steering Network with Scale Estimation

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F21%3A00357705" target="_blank" >RIV/68407700:21230/21:00357705 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1109/IROS51168.2021.9635912" target="_blank" >https://doi.org/10.1109/IROS51168.2021.9635912</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/IROS51168.2021.9635912" target="_blank" >10.1109/IROS51168.2021.9635912</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Monocular Teach-and-Repeat Navigation using a Deep Steering Network with Scale Estimation

Popis výsledku v původním jazyce

This paper proposes a novel monocular teach-and-repeat navigation system with the capability of scale awareness, i.e. the absolute distance between observation and goal images. It decomposes the navigation task into a sequence of visual servoing sub-tasks to approach consecutive goal/node images in a topological map. To be specific, a novel hybrid model, named deep steering network is proposed to infer the navigation primitives according to the learned local feature and scale for each visual servoing sub-task. A novel architecture, Scale-Transformer, is developed to estimate the absolute scale between the observation and goal image pair from a set of matched deep representations to assist repeating navigation. The experiments demonstrate that our scale-aware teach-and-repeat method achieves satisfying navigation accuracy, and converges faster than the monocular methods without scale correction given an inaccurate initial pose.

Název v anglickém jazyce

Monocular Teach-and-Repeat Navigation using a Deep Steering Network with Scale Estimation

Popis výsledku anglicky

This paper proposes a novel monocular teach-and-repeat navigation system with the capability of scale awareness, i.e. the absolute distance between observation and goal images. It decomposes the navigation task into a sequence of visual servoing sub-tasks to approach consecutive goal/node images in a topological map. To be specific, a novel hybrid model, named deep steering network is proposed to infer the navigation primitives according to the learned local feature and scale for each visual servoing sub-task. A novel architecture, Scale-Transformer, is developed to estimate the absolute scale between the observation and goal image pair from a set of matched deep representations to assist repeating navigation. The experiments demonstrate that our scale-aware teach-and-repeat method achieves satisfying navigation accuracy, and converges faster than the monocular methods without scale correction given an inaccurate initial pose.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/GC20-27034J" target="_blank" >GC20-27034J: Rozšíření prostorových modelů explicitní representací času pro dlouhodobou autonomii mobilních robotů</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2021

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

ISBN

978-1-6654-1714-3

ISSN

2153-0858

e-ISSN

2153-0866

Počet stran výsledku

7

Strana od-do

2613-2619

Název nakladatele

IEEE

Místo vydání

Piscataway

Místo konání akce

Praha

Datum konání akce

27. 9. 2021

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000755125502022