Filters








121,326 Hits in 4.7 sec

A Scenario Description Language Based on Action Frame

Hong Hui Zhang, Atsushi Ohnishi
2003 International Conference on Advanced Information Systems Engineering  
This paper describes a language for describing scenarios in which simple action traces are embellished to include typed frames based on a simple case grammar of actions.  ...  Scenarios are informal, and are difficult to be processed automatically.  ...  Scenarios can be regarded as a sequence of events. Each event corresponds to a certain action. We have developed a scenario description language based on this notion.  ... 
dblp:conf/caise/HuiO03 fatcat:dyrzpahak5dpzc2bidvrv2ji6a

PixL2R: Guiding Reinforcement Learning Using Natural Language by Mapping Pixels to Rewards [article]

Prasoon Goyal, Scott Niekum, Raymond J. Mooney
2020 arXiv   pre-print
In this work, we propose a model that directly maps pixels to rewards, given a free-form natural language description of the task, which can then be used for policy learning.  ...  Our experiments on the Meta-World robot manipulation domain show that language-based rewards significantly improves the sample efficiency of policy learning, both in sparse and dense reward settings.  ...  As such, a description collected for one scenario for a task can be paired with any of the 100 scenarios for the corresponding task.  ... 
arXiv:2007.15543v2 fatcat:5woqozb4zrdw7hqg7sruagrmiy

Spatial relation learning in complementary scenarios with deep neural networks

Jae Hee Lee, Yuan Yao, Ozan Özdemir, Mengdi Li, Cornelius Weber, Zhiyuan Liu, Stefan Wermter
2022 Frontiers in Neurorobotics  
First, using an embodied model, the agent learns to reach toward an object based on simple instructions involving left-right relations.  ...  and show that recent reasoning models can learn directional relations in different frames of reference.  ...  predefined language descriptions and simple robotic manipulation actions on objects.  ... 
doi:10.3389/fnbot.2022.844753 pmid:35966371 pmcid:PMC9366005 fatcat:mnkk5c3rizf6rnguc22eiszc3y

Interactive Planning of Manual Assembly Operations: From Language to Motion

Stephan Busemann, Jörg Steffen, Erik Herrmann
2016 Procedia CIRP  
We suggest solutions to the subtask of mapping textual descriptions onto motion sequences. Consider a description like "Tighten arm support with cordless screw driver on center console".  ...  Evaluating this kind of condition-action rules against constraints imposed by the scene for a given textual description leads to a sequence of elementary actions that is then processed further and visualized  ...  Breakdown of task descriptions The breakdown of a case frame into elementary actions depends on the state of the scenario the action shall be carried out on.  ... 
doi:10.1016/j.procir.2015.12.106 fatcat:u47m55c37nbt3dlmuf2kvxvjmy

A Generation Method of Exceptional Scenarios from a Normal Scenario

A. OHNISHI
2008 IEICE transactions on information and systems  
This paper proposes a method to generate exceptional scenarios from a normal scenario written with a scenario language.  ...  This method includes (1) generation of exceptional plans and (2) generation of exceptional scenario by a user's selection of these plans.  ...  In this scenario language, simple action traces are embellished including typed frames based on a simple case grammar of actions and for describing the sequence among events [12] , [19] .  ... 
doi:10.1093/ietisy/e91-d.4.881 fatcat:7pybcxkz4rfuhdlwefaijht6p4

An Open Vocabulary Semantic Parser for End-User Programming Using Natural Language

Juliano Efson Sales, Andr� Freitas, Siegfried Handschuh
2018 2018 IEEE 12th International Conference on Semantic Computing (ICSC)  
In this paper, we propose a semantic parsing approach to map natural language commands to actions from a large and heterogeneous frame set trained under a small set of annotated data.  ...  of a programming language.  ...  Let A be a knowledge base (KB) composed of a set of k action frames (a 1 , a 2 , . . . , a k ).  ... 
doi:10.1109/icsc.2018.00020 dblp:conf/semco/SalesFH18 fatcat:tf3eo5qfkvc3hp5y6tmg5gzmpe

SemEval-2017 Task 11: End-User Development using Natural Language

Juliano Sales, Siegfried Handschuh, André Freitas
2017 Proceedings of the 11th International Workshop on Semantic Evaluation (SemEval-2017)  
This task proposes a challenge to support the interaction between users and applications, micro-services and software APIs using natural language.  ...  scenarios of high lexical and semantic heterogeneity.  ...  Table 2 : 2 Examples of action frames used in the scenarios.  ... 
doi:10.18653/v1/s17-2092 dblp:conf/semeval/SalesHF17 fatcat:4e2cet6h7fh4dfkxmvbdg5vvem

Frame Interpretation and Validation in a Open Domain Dialogue System [article]

Artur Ventura, Nuno Diegues, David Martins de Matos
2012 arXiv   pre-print
Our goal in this paper is to establish a means for a dialogue platform to be able to cope with open domains considering the possible interaction between the embodied agent and humans.  ...  To this end we present an algorithm capable of processing natural language utterances and validate them against knowledge structures of an intelligent agent's mind.  ...  Also, each language possesses a grammar description: each verb on the language has an associated set of structural (NP VP NP) and semantic (Agent V Object) relations between itself and its arguments.  ... 
arXiv:1207.4307v1 fatcat:h745kndun5dalnhgqxhmyfgzhe

Enabling Robots to Understand Incomplete Natural Language Instructions Using Commonsense Reasoning [article]

Haonan Chen, Hao Tan, Alan Kuntz, Mohit Bansal, Ron Alterovitz
2020 arXiv   pre-print
perform tasks based on incomplete natural language instructions.  ...  However, natural language instructions are often missing information that would be obvious to a human based on environmental context and common sense, and hence does not need to be explicitly stated.  ...  In this way, we can construct a test scenario with k candidates, where one of them is plausible based on the human annotation.  ... 
arXiv:1904.12907v2 fatcat:gs4g2ijdnrds7afl6dggvbtwlu

The University of Sheffield and University of Engineering & Technology, Lahore at TRECVID 2016: Video to Text Description Task

Samyan Qayyum Wahla, Sahar Waqar, Muhammad Usman Ghani Khan, Yoshihiko Gotoh
2016 TREC Video Retrieval Evaluation  
Our approach is based on the extraction of high level features using image processing and computer vision stage, scene recognition using HLFs with machine learning approach and generation of natural language  ...  Second approach is totally based on facial features and gestures. Third approach is similar to first approach but machine learning approach is replaced with deep learning.  ...  Natural Language Generation Stage Action is recognized per 8 frames which mean ADR gives no output per 7 frames and on 8th frame sentence contains a verb.  ... 
dblp:conf/trecvid/WahlaWKG16 fatcat:k7ebopwdm5fghgsaj2e6uhvtze

Scenario-Based Modeling of Multi-Agent Systems [chapter]

Armin Stranjak, Igor Cavrak, Mario Zagar
2011 Multi-Agent Systems - Modeling, Interactions, Simulations and Case Studies  
on the agent's action definition in the scenario description(s).  ...  Metadata properties are defined on a per-language element basis (role, scenario, action).  ... 
doi:10.5772/15671 fatcat:lxupbc3ekjhvdkqoz6l6ok3czu

Speech Communication in Teaching a Foreign Language

2019 International journal of recent technology and engineering  
One of the main modules in the system of any developed language is the sublanguage of everyday communication.The article deals with interpersonal communication in teaching foreign languages.  ...  A language is considered not only and not so much as a system of organized levels, but primarily as a system of a number of sub-languages or modules, each of which, in turn, has a system organization.  ...  The concept is thus based on the concept of "language in action".  ... 
doi:10.35940/ijrte.c1101.1083s19 fatcat:b5findh5tzgahmvsp3gs442glm

Rule-based verification of scenarios with pre-conditions and post-conditions

T. Toyama, A. Ohnishi
2005 13th IEEE International Conference on Requirements Engineering (RE'05)  
The purposes are to include typed frames based on a simple case grammar of actions and to describe the sequence among events.  ...  Based on this scenario language, this paper describes both (1) a correctness-verification method using rules to detect errors (lack of events, extra events, and wrong sequence among events) in a scenario  ...  The authors have developed a scenario language for describing scenarios in which simple action traces are embellished to include typed frames based on a simple case grammar of actions and to describe the  ... 
doi:10.1109/re.2005.63 dblp:conf/re/ToyamaO05 fatcat:lsdfirwlkned3otafmdtxmcybi

Language is Power: Representing States Using Natural Language in Reinforcement Learning [article]

Erez Schwartz, Guy Tennenholtz, Chen Tessler, Shie Mannor
2020 arXiv   pre-print
Language can represent complex scenarios and concepts, making it a favorable candidate for representation.  ...  While previous methods focused on information in its raw form (e.g., raw visual input), we propose to represent the state using natural language.  ...  In each patch, objects were counted, and a sentence description was constructed based on this information.  ... 
arXiv:1910.02789v2 fatcat:nsoh4gxudzczrcyeblalirtqui

A framework for creating natural language descriptions of video streams

Muhammad Usman Ghani Khan, Nouf Al Harbi, Yoshihiko Gotoh
2015 Information Sciences  
These features are converted into natural language descriptions using a template-based approach built on a context free grammar, incorporating spatial and temporal information.  ...  Additionally a task-based scheme is introduced that provides quantitative evaluation for relevance of generated descriptions.  ...  Sample Scenario for Description Generation When creating natural language description of a scene, visual features are identified first.  ... 
doi:10.1016/j.ins.2014.12.034 fatcat:tjpyhsq5kzc4xcp4ooqsogdd5e
« Previous Showing results 1 — 15 out of 121,326 results