Feng, Jianxiang und Durner, Maximilian und Marton, Zoltan-Csaba und Balint-Benczedi, Ferenc und Triebel, Rudolph (2022) Introspective Robot Perception using Smoothed Predictions from Bayesian Neural Networks. In: 19th International Symposium of Robotics Research, ISRR 2019, 20. Springer, Cham. International Symposium on Robotics Research (ISRR) 2019, 06-10 Oct 2019, Hanoi, Vietnam. ISBN 978-3-030-95458-1.
Dies ist die aktuellste Version dieses Eintrags.
PDF
1MB |
Offizielle URL: https://link.springer.com/chapter/10.1007/978-3-030-95459-8_40
Kurzfassung
This work focuses on improving uncertainty estimation in the field of object classification from RGB images and demonstrates its benefits in two robotic applications. We employ a Bayesian Neural Network (BNN), and evaluate two practical inference techniques to obtain better uncertainty estimates, namely Concrete Dropout (CDP) and Kronecker-factored Laplace Approximation (LAP). We show a performance increase using more reliable uncertainty estimates as unary potentials within a Conditional Random Field (CRF), which is able to incorporate contextual information as well. Furthermore, the obtained uncertainties are exploited to achieve domain adaptation in a semi-supervised manner, which requires less manual efforts of annotating data. We evaluate our approach on two public benchmark datasets that are relevant for robot perception tasks.
elib-URL des Eintrags: | https://elib.dlr.de/129230/ | ||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Dokumentart: | Konferenzbeitrag (Vortrag) | ||||||||||||||||||||||||
Titel: | Introspective Robot Perception using Smoothed Predictions from Bayesian Neural Networks | ||||||||||||||||||||||||
Autoren: |
| ||||||||||||||||||||||||
Datum: | 17 Februar 2022 | ||||||||||||||||||||||||
Erschienen in: | 19th International Symposium of Robotics Research, ISRR 2019 | ||||||||||||||||||||||||
Referierte Publikation: | Ja | ||||||||||||||||||||||||
Open Access: | Ja | ||||||||||||||||||||||||
Gold Open Access: | Nein | ||||||||||||||||||||||||
In SCOPUS: | Nein | ||||||||||||||||||||||||
In ISI Web of Science: | Nein | ||||||||||||||||||||||||
Band: | 20 | ||||||||||||||||||||||||
Verlag: | Springer, Cham | ||||||||||||||||||||||||
Name der Reihe: | Springer Proceedings in Advanced Robotics | ||||||||||||||||||||||||
ISBN: | 978-3-030-95458-1 | ||||||||||||||||||||||||
Status: | veröffentlicht | ||||||||||||||||||||||||
Stichwörter: | BNN, CRF, introspective classification | ||||||||||||||||||||||||
Veranstaltungstitel: | International Symposium on Robotics Research (ISRR) 2019 | ||||||||||||||||||||||||
Veranstaltungsort: | Hanoi, Vietnam | ||||||||||||||||||||||||
Veranstaltungsart: | internationale Konferenz | ||||||||||||||||||||||||
Veranstaltungsdatum: | 06-10 Oct 2019 | ||||||||||||||||||||||||
Veranstalter : | IFRR (International Foundation of Robotics Research) | ||||||||||||||||||||||||
HGF - Forschungsbereich: | Luftfahrt, Raumfahrt und Verkehr | ||||||||||||||||||||||||
HGF - Programm: | Raumfahrt | ||||||||||||||||||||||||
HGF - Programmthema: | Robotik | ||||||||||||||||||||||||
DLR - Schwerpunkt: | Raumfahrt | ||||||||||||||||||||||||
DLR - Forschungsgebiet: | R RO - Robotik | ||||||||||||||||||||||||
DLR - Teilgebiet (Projekt, Vorhaben): | R - Multisensorielle Weltmodellierung (RM) [RO] | ||||||||||||||||||||||||
Standort: | Oberpfaffenhofen | ||||||||||||||||||||||||
Institute & Einrichtungen: | Institut für Robotik und Mechatronik (ab 2013) > Perzeption und Kognition | ||||||||||||||||||||||||
Hinterlegt von: | Feng, Jianxiang | ||||||||||||||||||||||||
Hinterlegt am: | 23 Sep 2019 09:25 | ||||||||||||||||||||||||
Letzte Änderung: | 27 Sep 2022 11:30 |
Verfügbare Versionen dieses Eintrags
- Introspective Robot Perception using Smoothed Predictions from Bayesian Neural Networks. (deposited 23 Sep 2019 09:25) [Gegenwärtig angezeigt]
Nur für Mitarbeiter des Archivs: Kontrollseite des Eintrags