changed 3 years ago  
Published Linked with GitHub

A Short Introduction to Spoken Language Processing

Dr. Chen-Yu Chiang (江振宇博士)
Speech & Multimedia Signal Processing Lab (SMSPLab)
Dept. of Communication Engineering,
National Taipei University, Taiwan

Spoken Language Processing is Ubiquitous! (1/3)

  • Human to human phone communications:
    • local call (家用電話)
    • mobile phone (行動電話)
    • voice over internet protocol (VoIP like Skype)
    • concall/video/audio conferece (Google Meet/Webex/Team/Zoom/jitsi)

Spoken Language Processing is Ubiquitous! (2/3)

  • Human-machine communication:
    • Apple Siri
    • Amazon Alexa
    • Google home
    • 小米音箱
    • Google voice search

Spoken Language Processing is Ubiquitous! (3/3)

  • Assistive systems:
    • speech-generating device
    • speech-to-speech translation (e.g., Mandarin <–> English)
    • computer-assisted language learning system
    • talkback (語音輔助)
    • 雅婷逐字稿
    • Any device that generates speech to inform you something!

Spoken Language Processing Technologies

  • Speech coding and compression
  • Speech recognition (speech to text)
  • Text to speech (TTS)/speech synthesis
  • Voice conversion
  • Spoken dialog system
If we do not have speech coding and compression by machines …

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →

source: https://www.youtube.com/watch?v=zQHhbhtpJ3M

Speech to speech translation

Voice conversion

Sampling/cloning your voice!

  • 在DJ雷欧的家中,洛托姆图鉴启动了新功能扩展——声音取样功能,同时洛托姆图鉴的假发也被库库伊博士送给了地鼠。

  • 在阿羅拉!第一次的教學觀摩!!中,洛托姆图鉴运用声音取样功能,在教学观摩活动上代替小智进行发言,但结果由于洛托姆图鉴的语速越来越快以及突然用成也・大木的声音说出了宝可梦谐音冷笑话并且说出了“洛托”的口头禅而穿帮。

source: http://www.linban.com/j/197604.shtml and https://wiki.52poke.com/wiki/宝可梦_太阳＆月亮_第24集
Stephen Hawking's speech-generating device, 2008

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →

source: https://www.youtube.com/watch?v=xjBIsp8mS-c&t=508s
Project Revoice: What it means to lose your ability to speak (2018)

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →

source: https://www.youtube.com/watch?v=6A2zt3JrPFo&t=0s
Interview with Dr. Peter Scott-Morgan [Peter 2.0]: Changing what it means to be human (2022)

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →

source: https://www.youtube.com/watch?v=8FkxDYvDL7U&t=195s
Meet The 'Human Cyborg' Defying Motor Neuron Disease With The Help Of Technology (2021)

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →

source: https://www.youtube.com/watch?v=WQdCxQ6rG5s

Scientist Attempts To Overcome MND! | Peter: The Human Cyborg (2021)

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →

source: https://www.youtube.com/watch?v=u3ovqznXZLs
回聲計畫/重 聲 save & sound (2020-2021)

Image Not Showing Possible Reasons
  • The image file may be corrupted
  • The server hosting the image is unavailable
  • The image path is incorrect
  • The image format is not supported
Learn More →

source: https://www.youtube.com/watch?v=XYeAfpCGwIk

End

© Speech & Multimedia Signal Processing Lab (SMSPLab), National Taipei University, New Taipei City, Taiwan, 2012-2022

Select a repo