A Human Voice Song Requesting System Based on Connected Vehicle in Cloud Computing

  • Ding Yi
  • Jian Zhang
Conference paper
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 260)


Traditional on-board vehicle computer (OVC) systems do not support speech recognition and do not have massive song library due to their limited storage capacity and computing power. Furthermore, it is very inconvenient to update the song library to keep it up-to-date. The advent of cloud computing in conjunction with mobile computing has enabled human voice song requesting and mass song playing from on-board computer in a vehicle. A system of human voice song requesting from OVC connected to the music cloud has been proposed. Self-adaptive speech recognition and speech recognition technology based on keywords from song titles has been adopted. By recognizing the voice on a song title in a cloud environment, the digital content of the selected song stored in the cloud is streamed to and then played back in the OVC. As a result, the system improves the safety of driving and enables automobile entertainment more humane.


Networking of vehicles On-board vehicle computer (OVC) Speech recognition Human voice song requesting (HVSR) Cloud computing Music cloud 


  1. 1.
    Wang Z (2004) Research in the application of streaming media technology. Comput Eng 5:34–36Google Scholar
  2. 2.
    Chen X (2005) The design of embedded streaming media player. Zhejiang University, pp 17–19Google Scholar
  3. 3.
    Zhao B, Yan F, Zhang L, Wang J (2012) The construction of dependable cloud computing environment. China Comput Fed Commun 8(7):28Google Scholar
  4. 4.
  5. 5.
    Microsoft Tellme Speech Innovation:
  6. 6.
    Li N, Xu S, Ma Z, Shi L (2011) Simulation research on adaptive speech recognition algorithm. Comput Simul 8:181–185Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2014

Authors and Affiliations

  1. 1.School of Computer EngineeringShenzhen PolytechnicShenzhenChina

Personalised recommendations