深度学习技术在复杂信息保障中的应用现状与趋势
Current Situation and Future Development of Deep Learning Technology in Complex Information Support
邵振峰 1张红萍 2吴长枝 3齐晓飞 3黄俊 4汪家明1
作者信息
- 1. 武汉大学测绘遥感信息工程国家重点实验室,湖北 武汉,430079
- 2. 武汉大学测绘遥感信息工程国家重点实验室,湖北 武汉,430079;海军工程大学电子工程学院,湖北 武汉,430033
- 3. 西安测绘研究所,陕西 西安,710054
- 4. 陆军装备部驻武汉地区第二军事代表室,湖北 武汉,430000
- 折叠
摘要
随着人工智能技术和理论的发展和完善,各种算法模型的实用性和性能也逐渐提高,其必将给未来复杂多变的信息保障应用带来颠覆性变革.本文分析了基于深度学习的遥感影像融合、目标检测、数字孪生、网络攻防以及图像文字的篡改等技术,探索如何有效挖掘物理区域和信息环境中的多源异构信息用于辅助决策或者信息安全保障任务,构建智能的复杂作战信息保障系统.同时,提升软/硬件不可控性和算法的可解释性以及发展微型集群化和保护强度化的设备都是智能化信息保障理论方向的主要任务.
Abstract
With the development and improvement of artificial intelligence,the practicality and performance of various algo-rithm models are gradually improved,which will bring disrup-tive changes to the complex and changeable information sup-port and intelligent militarization application in the future.The paper analyzes the technologies of remote sensing image fusion,target detection,digital twinning,network attack,network defense,and image/text tampering based on deep learning.Then it explore how to effectively mine the multi-source heterogeneous information in the physical combat area and information environment for decision-making or informa-tion security support tasks,and builds an intelligent informa-tion support system.At the same time,improving the uncon-trollability of software/hardware and the interpretability of al-gorithms,as well as developing micro clustering and protec-tion intensive equipment are the main tasks of the theoretical direction of intelligent information support.
关键词
人工智能/信息保障/智能感知/深度伪造Key words
artificial intelligence/battlefield information sup-port/intellisense/intelligence deepfake引用本文复制引用
出版年
2024
NETL
NSTL
万方数据