EUVIMEDEUVIMED entdecken
Crossref

Navigating to objects in the real world

Theophile Gervet iD, Soumith Chintala, Dhruv Batra iD, Jitendra Malik iD, Devendra Singh Chaplot

Science Robotics · 28.06.2023 · 8 (79)

DOI: 10.1126/scirobotics.adf6991 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Science Robotics, erschienen am 28.06.2023. Verfasst von Theophile Gervet und 4 weiteren Autor:innen. Beteiligte Einrichtung: Carnegie Mellon University, Pittsburgh, PA, USA. Die Arbeit wurde bisher 129-mal zitiert (Crossref) und nennt 117 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Semantic navigation is necessary to deploy mobile robots in uncontrolled environments such as homes or hospitals. Many learning-based approaches have been proposed in response to the lack of semantic understanding of the classical pipeline for spatial navigation, which builds a geometric map using depth sensors and plans to reach point goals. Broadly, end-to-end learning approaches reactively map sensor inputs to actions with deep neural networks, whereas modular learning approaches enrich the classical pipeline with learning-based semantic sensing and exploration. However, learned visual navigation policies have predominantly been evaluated in sim, with little known about what works on a robot. We present a large-scale empirical study of semantic visual navigation methods comparing representative methods with classical, modular, and end-to-end learning approaches across six homes with no prior experience, maps, or instrumentation. We found that modular learning works well in the real world, attaining a 90% success rate. In contrast, end-to-end learning does not, dropping from 77% sim to a 23% real-world success rate because of a large image domain gap between sim and reality. For practitioners, we show that modular learning is a reliable approach to navigate to objects: Modularity and abstraction in policy design enable sim-to-real transfer. For researchers, we identify two key issues that prevent today’s simulators from being reliable evaluation benchmarks—a large sim-to-real gap in images and a disconnect between sim and real-world error modes—and propose concrete steps forward.

Autor:innen und Institutionen

  • Carnegie Mellon University, Pittsburgh, PA, USA.
  • Meta AI Research, Menlo Park, CA, USA.
  • Georgia Institute of Technology, Atlanta, GA, USA.
  • University of California, Berkeley, CA, USA.

ORCID-Profile: Theophile Gervet, Dhruv Batra, Jitendra Malik

Zitieren

APA

Gervet, T., Chintala, S., Batra, D., Malik, J., & Chaplot, D. S. (2023). Navigating to objects in the real world. Science Robotics, 8(79). https://doi.org/10.1126/scirobotics.adf6991

Vancouver

Gervet T, Chintala S, Batra D, Malik J, Chaplot DS. Navigating to objects in the real world. Science Robotics. 2023;8(79). doi:10.1126/scirobotics.adf6991

BibTeX
@article{Gervet2023,
  author = {Gervet, Theophile and Chintala, Soumith and Batra, Dhruv and Malik, Jitendra and Chaplot, Devendra Singh},
  title = {Navigating to objects in the real world},
  journal = {Science Robotics},
  year = {2023},
  volume = {8},
  number = {79},
  doi = {10.1126/scirobotics.adf6991}
}
Kostenlos starten →