MazeExplorer:用于评估强化学习泛化的可定制3D基准

2019 IEEE Conference on Games (CoG) Pub Date : 2019-12-05 DOI:10.1109/CIG.2019.8848048

Luke Harries, Sebastian Lee, Jaroslaw Rzepecki, Katja Hofmann, Sam Devlin

{"title":"MazeExplorer:用于评估强化学习泛化的可定制3D基准","authors":"Luke Harries, Sebastian Lee, Jaroslaw Rzepecki, Katja Hofmann, Sam Devlin","doi":"10.1109/CIG.2019.8848048","DOIUrl":null,"url":null,"abstract":"This paper presents a customisable 3D benchmark for assessing generalisability of reinforcement learning agents based on the 3D first-person game Doom and open source environment VizDoom. As a sample use-case we show that different domain randomisation techniques during training in a key-collection navigation task can help to improve agent performance on unseen evaluation maps.","PeriodicalId":177208,"journal":{"name":"2019 IEEE Conference on Games (CoG)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"10","resultStr":"{\"title\":\"MazeExplorer: A Customisable 3D Benchmark for Assessing Generalisation in Reinforcement Learning\",\"authors\":\"Luke Harries, Sebastian Lee, Jaroslaw Rzepecki, Katja Hofmann, Sam Devlin\",\"doi\":\"10.1109/CIG.2019.8848048\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This paper presents a customisable 3D benchmark for assessing generalisability of reinforcement learning agents based on the 3D first-person game Doom and open source environment VizDoom. As a sample use-case we show that different domain randomisation techniques during training in a key-collection navigation task can help to improve agent performance on unseen evaluation maps.\",\"PeriodicalId\":177208,\"journal\":{\"name\":\"2019 IEEE Conference on Games (CoG)\",\"volume\":\"1 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2019-12-05\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"10\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2019 IEEE Conference on Games (CoG)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/CIG.2019.8848048\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 IEEE Conference on Games (CoG)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/CIG.2019.8848048","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 10

摘要

本文提出了一个可定制的3D基准，用于评估基于3D第一人称游戏《毁灭战士》和开源环境VizDoom的强化学习代理的通用性。作为一个示例用例，我们展示了在键集导航任务的训练过程中不同的领域随机化技术可以帮助提高智能体在看不见的评估地图上的性能。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

MazeExplorer: A Customisable 3D Benchmark for Assessing Generalisation in Reinforcement Learning

This paper presents a customisable 3D benchmark for assessing generalisability of reinforcement learning agents based on the 3D first-person game Doom and open source environment VizDoom. As a sample use-case we show that different domain randomisation techniques during training in a key-collection navigation task can help to improve agent performance on unseen evaluation maps.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

2019 IEEE Conference on Games (CoG)

自引率

0.00%

发文量