您好,欢迎来到一带一路数据库!

全库
全文
  • 全文
  • 标题
  • 所属丛书
  • 作者/机构
  • 关键词
  • 主题词
  • 摘要
高级检索

您好,欢迎来到一带一路数据库!

开源框架成为科技巨头全面布局的重点

作者:张宇泽 出版日期:2017年06月 报告页数:17 页 报告大小: 报告字数:12285 字 所属丛书:工业和信息化蓝皮书 所属图书:人工智能发展报告(2016~2017) 浏览人数: 下载人数:

文章摘要:开源深度学习平台是近几年推动人工智能技术发展的重要动力。开源深度学习平台优点显著:它允许公众使用、复制、修改源代码。开放共享的灵魂内核与互联网类似,相对于传统闭源软件平台,开源平台汇聚更广泛的集体智慧,具有更新速度快、拓展性强等特点,是人工智能发展强有力的助推器,并且大幅降低软件企业开发成本和客户的购买成本。深度学习作为人工智能领域最核心的技术,提供了传统方法无法比拟的优势。由于深度学习的开发和部署涉及编程语言、接口、操作系统、CPU、G... 展开

文章摘要:开源深度学习平台是近几年推动人工智能技术发展的重要动力。开源深度学习平台优点显著:它允许公众使用、复制、修改源代码。开放共享的灵魂内核与互联网类似,相对于传统闭源软件平台,开源平台汇聚更广泛的集体智慧,具有更新速度快、拓展性强等特点,是人工智能发展强有力的助推器,并且大幅降低软件企业开发成本和客户的购买成本。深度学习作为人工智能领域最核心的技术,提供了传统方法无法比拟的优势。由于深度学习的开发和部署涉及编程语言、接口、操作系统、CPU、GPU等纷繁复杂的软硬件平台,因此需要框架提供高层的操作接口,从而让使用者更聚焦于编程运行而无须关注底层细节。早期深度学习框架多为学术机构提供,后来随着人工智能竞争激烈化,各大科技巨头力推自家开源框架,建立自家深度学习生态体系。

收起

Abstract:Open-source deep learning platform is the key of artificial intelligent(AI)development in past few years. It has significant advantages such as it allows public access,duplication and modification of source code,the opened and shared kernel is similar to internet. As compared to the traditional closed-source,open-source is developed in a collaborative public manner,with the feature of rapid evolution and more scalable thus a boo... 展开

Abstract:Open-source deep learning platform is the key of artificial intelligent(AI)development in past few years. It has significant advantages such as it allows public access,duplication and modification of source code,the opened and shared kernel is similar to internet. As compared to the traditional closed-source,open-source is developed in a collaborative public manner,with the feature of rapid evolution and more scalable thus a booster for AI development. In addition,it tremendously reduces company’s development cost,hence a lower consumer prices. As the core technology in AI,deep learning outranks the old technology. This is because deep learning requires various software and hardware platforms such as programming languages,interfaces,OS,CPU,GPU et cetera,thus a demand for high level interfaces from the frame which allows the users to focus on programming but not low level details. On the other hand,deep learning frame in earlier stage was mainly provided by academic institutions which is not comparative for today’s AI. As a result,tech giants start to develop their own open-source frame and deep learning ecosystem.

收起

作者简介

张宇泽:张宇泽,工学硕士,毕业于中国科学院,主要跟踪国内外智能语音、人工智能等多个领域企业、战略规划和产业的发展动向,在智能语音及人工智能领域具有丰富的研究经验。