手机终端人工智能关键技术研究

庞涛; 丘海华; 潘碧莹

doi:10.11959/j.issn.1000-0801.2020045

您当前的位置：

首页 >

文章列表页 >

手机终端人工智能关键技术研究

运营技术广角 | 更新时间：2024-06-05

- 手机终端人工智能关键技术研究
- Research on artificial intelligence key technologies of mobile terminal
- 电信科学 2020年36卷第5期页码：145-151
- 作者机构：
- 作者简介：
  
  [ "庞涛（1977- ），男，中国电信股份有限公司研究院工程师，主要研究方向为大数据应用、人工智能、智能终端技术等" ]
  [ "丘海华（1991- ），男，中国电信股份有限公司研究院工程师，主要研究方向为人工智能、AR技术、边缘计算等" ]
  [ "潘碧莹（1994- ），女，中国电信股份有限公司研究院工程师，主要研究方向为人工智能、智能终端技术等" ]
- 基金信息：
- DOI：10.11959/j.issn.1000-0801.2020045
  中图分类号： TP334.1
- 网络出版日期：2020-05，
  
  纸质出版日期：2020-05-20
- 稿件说明：
移动端阅览
庞涛, 丘海华, 潘碧莹. 手机终端人工智能关键技术研究[J]. 电信科学, 2020,36(5):145-151.

Tao PANG, Haihua QIU, Biying PAN. Research on artificial intelligence key technologies of mobile terminal[J]. Telecommunications science, 2020, 36(5): 145-151.
庞涛, 丘海华, 潘碧莹. 手机终端人工智能关键技术研究[J]. 电信科学, 2020,36(5):145-151. DOI： 10.11959/j.issn.1000-0801.2020045.

Tao PANG, Haihua QIU, Biying PAN. Research on artificial intelligence key technologies of mobile terminal[J]. Telecommunications science, 2020, 36(5): 145-151. DOI： 10.11959/j.issn.1000-0801.2020045.

摘要

随着移动终端软硬件技术的进步，移动终端的机器学习能力得以挖掘。从手机终端人工智能技术整体框架入手，研究了端侧人工智能硬件加速技术，对比了当前主流端侧机器学习框架和神经网络模型压缩等软件技术，分析了AI应用发展趋势，总结了手机终端人工智能技术进展和未来发展趋势。

Abstract

With the progress of software and hardware technology of mobile terminals

the machine learning ability of mobile terminals has been explored.Starting from the overall framework of mobile terminal artificial intelligence technology

the end-to-end artificial intelligence hardware acceleration technology was studied

mainstream on-device machine learning frameworks

neural network model compression and other software technologies were compared

the development trend of AI application was analyzed

and the current technological progress and future development trend of mobile terminal artificial intelligence were summarized.

关键词

Keywords

references

NEIL C , THOMPSON , SVENJA SPANUTH . The decline of computers as a general purpose technology:why deep learning and the end of Moore’s law are fragmenting computing [J ] . 2018 .

艾瑞咨询 . 中国人工智能手机行业研究报告 [R ] . 2018 .

IResearch . China artificial intelligence mobile phone industry research report [R ] . 2018 .

中国电信 . 中国电信全网通AI手机白皮书2.0 [R ] . 2018 .

China Telecom . China Telecom full netcom AI mobile white paper 2.0 [R ] . 2018 .

CHEN Y J , CHEN T S , XU Z W , et al . DianNao family:energy-efficient hardware accelerators for machine learning [J ] . Communications of the ACM , 2016 , 59 ( 11 ): 105 - 112 .

CHENGJ , WANG PS , LIG , et al . Recent advances in efficient computation of deep convolutional neural networks [J ] . arXiv,2018:1802.00939 ,

LI H , ASIM KADAV , IGOR DURDANOVIC , et al . Pruning filters for efficient convnets [J ] .arXiv,ICLR 2017:1608.08710.

HAN S , MAO H , DALLY W J . Deep compression:compressing deep neural networks with pruning,trained quantization and Huffman coding [J ] . arXiv Preprint arXiv , 2015 :1510.00149.

BENOIT J , SKIRMANTAS K , BO C , et al . Quantization and training of neural networks for efficient integer-arithmetic-only inference [J ] . arXiv,2017:1712.05877 ,

SHENG T , FENG C , ZHUO S J , et al . A quantization-friendly separable convolution for mobilenets [J ] . arXiv , 2019 :1803.08607.

庞涛 . 开源深度学习框架发展现状与趋势研究 [J ] . 互联网天地 , 2018 ( 4 ): 46 - 54 .

PANG T . Research on development status and trend of open source deep learning framework [J ] . Internet World , 2018 ( 4 ): 46 - 54 .

余霆嵩 . 纵览轻量化卷积神经网络:SqueezeNet、MobileNet、ShuffleNet、Xception [J ] . 2018 .

YU T S . Overview of lightweight convolutional neural networks:SqueezeNet、MobileNet、ShuffleNet、Xception [J ] . 2018 .

HARD A , RAO K , MATHEWS R , et al . Federated learning for mobile keyboard prediction [J ] . arXiv,2018:1811.03604v1 ,

中国人工智能产业发展联盟（AIIA） . 手机人工智能技术与应用白皮书 [R ] . 2019 .

China Artificial Intelligence Industry Development Alliance (AIIA) . Mobile phone artificial intelligence technology and application white paper [R ] . 2019 .

浏览量

744

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

云化的智算中心万卡集群创新与实践

5G-Advanced核心网运行智能架构演进、关键技术及应用研究

Prophet模型在话单量智能预测中的应用研究

构建未来：通信大数据开放平台在智能社会治理中的创新应用与发展前景

基于可解释机器学习模型的电信行业客户流失预测研究