國立虎尾科技大學 |

Learning and Transferring Implicit Knowledge of Object Properties Across Robots via Interactive Perception.

Record Type:	Language materials, manuscript : Monograph/item
Title/Author:	Learning and Transferring Implicit Knowledge of Object Properties Across Robots via Interactive Perception./
Author:	Tatiya, Gyan.
Description:	1 online resource (232 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 85-08, Section: B.
Contained By:	Dissertations Abstracts International85-08B.
Subject:	Computer science. -
Online resource:	click for full text (PQDT)
ISBN:	9798381701555

Learning and Transferring Implicit Knowledge of Object Properties Across Robots via Interactive Perception.
Tatiya, Gyan.

Learning and Transferring Implicit Knowledge of Object Properties Across Robots via Interactive Perception. - 1 online resource (232 pages)

Source: Dissertations Abstracts International, Volume: 85-08, Section: B.

Thesis (Ph.D.)--Tufts University, 2024.

Includes bibliographical references

Integrating personal and service robots into real-world environments necessitates a deep understanding of complex object properties beyond visual attributes. The field of robotics has showcased the capabilities of robots in executing interactive perception tasks involving physical interactions with objects and the acquisition of implicit object properties through non-visual sensory signals such as audio, tactile, and haptic feedback. While learning-based approaches for interactive perception tasks have yielded success, a critical limitation is the time-intensive nature of individual robot learning, impeding the deployment of algorithms across large collections of robots with subtly different morphologies, as in warehouse or factory automation settings. To address this challenge, this dissertation introduces innovative frameworks, both theoretically and practically, facilitating the transfer of multi-sensory object representations among robots, ensuring that newly developed robots can build upon existing knowledge rather than starting from scratch. These frameworks encompass generating features for newly deployed robots by leveraging insights from experienced robots, developing shared latent feature spaces among robots, and acquiring unified multi-sensory object property representations that are transferable across different tasks. This dissertation contributes to advancing robot perception capabilities by enabling robots to share their perceptual knowledge and publishes three large object exploration datasets by robots to facilitate further investigations. Future work should extend the research horizon by delving into autonomous learning mechanisms for exploratory behaviors, refining adaptability with learning-based policies for handling complex tasks, and automating object selection algorithms to enhance the efficiency of perceptual knowledge transfer models across diverse robots.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798381701555Subjects--Topical Terms:

573171
Computer science.
Subjects--Index Terms:

Robot learningIndex Terms--Genre/Form:

554714
Electronic books.

Learning and Transferring Implicit Knowledge of Object Properties Across Robots via Interactive Perception.
LDR:03351ntm a22004097 4500 001 1142896
005 20240429063859.5
006 m o d
007 cr mn ---uuuuu
008 250605s2024 xx obm 000 0 eng d
020 $a 9798381701555
035 $a (MiAaPQ)AAI30988509
035 $a AAI30988509
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Tatiya, Gyan. $3 1467418
245 1 0 $a Learning and Transferring Implicit Knowledge of Object Properties Across Robots via Interactive Perception.
264 0 $c 2024
300 $a 1 online resource (232 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-08, Section: B.
500 $a Advisor: Sinapov, Jivko.
502 $a Thesis (Ph.D.)--Tufts University, 2024.
504 $a Includes bibliographical references
520 $a Integrating personal and service robots into real-world environments necessitates a deep understanding of complex object properties beyond visual attributes. The field of robotics has showcased the capabilities of robots in executing interactive perception tasks involving physical interactions with objects and the acquisition of implicit object properties through non-visual sensory signals such as audio, tactile, and haptic feedback. While learning-based approaches for interactive perception tasks have yielded success, a critical limitation is the time-intensive nature of individual robot learning, impeding the deployment of algorithms across large collections of robots with subtly different morphologies, as in warehouse or factory automation settings. To address this challenge, this dissertation introduces innovative frameworks, both theoretically and practically, facilitating the transfer of multi-sensory object representations among robots, ensuring that newly developed robots can build upon existing knowledge rather than starting from scratch. These frameworks encompass generating features for newly deployed robots by leveraging insights from experienced robots, developing shared latent feature spaces among robots, and acquiring unified multi-sensory object property representations that are transferable across different tasks. This dissertation contributes to advancing robot perception capabilities by enabling robots to share their perceptual knowledge and publishes three large object exploration datasets by robots to facilitate further investigations. Future work should extend the research horizon by delving into autonomous learning mechanisms for exploratory behaviors, refining adaptability with learning-based policies for handling complex tasks, and automating object selection algorithms to enhance the efficiency of perceptual knowledge transfer models across diverse robots.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Computer science. $3 573171
650 4 $a Robotics. $3 561941
650 4 $a Information technology. $3 559429
653 $a Robot learning
653 $a Multi-modal perception
653 $a Representation learning
653 $a Sensorimotor learning
653 $a Transfer learning
655 7 $a Electronic books. $2 local $3 554714
690 $a 0984
690 $a 0800
690 $a 0771
690 $a 0489
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a Tufts University. $b Computer Science. $3 1188601
773 0 $t Dissertations Abstracts International $g 85-08B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30988509 $z click for full text (PQDT)