2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2019
DOI: 10.1109/iros40897.2019.8967639
|View full text |Cite
|
Sign up to set email alerts
|

Word2vec to behavior: morphology facilitates the grounding of language in machines

Abstract: Fig. 1. Overview of the method. A: The initial values of a robot control policy's hidden layer is set by supplying the word2vec embedding associated with a command such as 'stop' to one neuron in the input layer. The policy is then downloaded on to a robot, and the sensor data generated by its movement is supplied to the remainder of the input layer (dotted arrow), further altering the hidden-and motor layers. After evaluation, the robot's behavior is scored against an objective function paired with the comman… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

0
5
0

Year Published

2020
2020
2023
2023

Publication Types

Select...
4
2

Relationship

0
6

Authors

Journals

citations
Cited by 7 publications
(5 citation statements)
references
References 27 publications
0
5
0
Order By: Relevance
“…Source code is available in the GitHub repository ( https://github.com/robodiff/robodiff ) ( 51 ). All other data are included in the manuscript and/or supporting information .…”
Section: Data Materials and Software Availabilitymentioning
confidence: 99%
“…Source code is available in the GitHub repository ( https://github.com/robodiff/robodiff ) ( 51 ). All other data are included in the manuscript and/or supporting information .…”
Section: Data Materials and Software Availabilitymentioning
confidence: 99%
“…Some studies generated actions from descriptions represented by pre-trained word embeddings [2][3] [6][11] [16]. Zhong et al [11], Matthews et al [3], and Lunch et al [2], in particular, generated actions from commands including unseen words. Zhong et al [11] used averaged pre-trained word embeddings for this task; however, this approach fails to generate actions from descriptions whose meaning depends on the order of the words.…”
Section: A Action Generation Using Pre-trained Word Embeddingsmentioning
confidence: 99%
“…Zhong et al [11] used averaged pre-trained word embeddings for this task; however, this approach fails to generate actions from descriptions whose meaning depends on the order of the words. Although Matthews et al [3] generated actions from single-word commands, the correspondence between a command and the corresponding action is fixed.…”
Section: A Action Generation Using Pre-trained Word Embeddingsmentioning
confidence: 99%
See 1 more Smart Citation
“…Several joint learning methods for language and motion have been proposed to address these issues. Using contextindependent word embeddings derived from a large corpus, [2] and [3] dealt with unseen words that were not included in the training data. In other studies, stepwise learning methods were proposed to integrate multiple models from different domains [4] [5].…”
Section: Introductionmentioning
confidence: 99%