Object-sensitive Deep Reinforcement Learning

Author:

Li Yuezhang,Sycara Katia,Iyer Rahul

Abstract

Deep reinforcement learning has become popular over recent years, showing superiority on different visual-input tasks such as playing Atari games and robot navigation. Although objects are important image elements, few work considers enhancing deep reinforcement learning with object characteristics. In this paper, we propose a novel method that can incorporate object recognition processing to deep reinforcement learning models. This approach can be adapted to any existing deep reinforcement learning frameworks. State-of-the-art results are shown in experiments on Atari games. We also propose a new approach called “object saliency maps” to visually explain the actions made by deep reinforcement learning agents.

Publisher

EasyChair

Cited by 28 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. PAROLE: Profitable Arbitrage in Optimistic Rollup with ERC-721 Token Transactions;2024 54th Annual IEEE/IFIP International Conference on Dependable Systems and Networks (DSN);2024-06-24

2. Learning for Semantic Knowledge Base-Guided Online Feature Transmission in Dynamic Channels;ICC 2024 - IEEE International Conference on Communications;2024-06-09

3. Integrating Expert Knowledge into Fuzzy Reinforcement Learning;2024 IEEE 18th International Symposium on Applied Computational Intelligence and Informatics (SACI);2024-05-23

4. Optimizing Edge Computing for Telemedicine: A Deep Reinforcement Learning Approach to Task Offloading and Resource Allocation;2024 6th International Conference on Communications, Information System and Computer Engineering (CISCE);2024-05-10

5. Non-Causal Control For Wave Energy Conversion Based on the Double Deep Q Network;2024 UKACC 14th International Conference on Control (CONTROL);2024-04-10

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3