Signal Processing

Speech Signal processing transforms speech acoustics to features which represent speech information (for instance the distribution of energy over time and frequency), and vice versa. 

• ·        Lectures 11-20 in Roger Moore’s course on Speech Processing  cover speech signal processing. They are written for Computer Science students and assume some mathematical knowledge. They make use of the visual programming language Pure Data.
  • Lecture 11 Signals [Slides Audio]
  • Lecture 12 Frequency Analysis [Slides Audio]
  • Lecture 13 Digital Signals [Slides Audio]
  • Lecture 14 Waveform Processing [Slides Audio]
  • Lecture 15 Fourier Transforms [Slides Audio]
  • Lecture 16 The Z Transform [Slides Audio]
  • Lecture 17 Linear Filters [Slides Audio]
  • Lecture 18 Linear Prediction [Slides Audio]
  • Lecture 19 Cepstral Analysis [Slides Audio]
  • Lecture 20 Final Overview [Slides Audio]
• ·        Dan Ellis  covers speech signal processing from first principles: this is a non-mathematical treatment.
• ·        A series of 6 lectures by Professor E. Ambikairajah on YouTube  covers Speech and Audio Processing, following Rabiner’s classic textbook. The urls are

https://www.youtube.com/watch?v=Xjzm7S__kBU
https://www.youtube.com/watch?v=Y_mSQ7tTlvQ
https://www.youtube.com/watch?v=lWH-Oh5KnNY
https://www.youtube.com/watch?v=nRhkzeeMqVk
https://www.youtube.com/watch?v=LwWKTTzcg_M
https://www.youtube.com/watch?v=H9qNuyD7xdg

• ·        Lecture 2 in the Columbia Speech Recognition course  is a more advanced treatment.
• ·        Hynek Hermansky’s take on speech signal processing is in part2 of his ISCA lectures.
• Speech Signal Processing or 'Feature Engineering'  is covered in Edinburgh's speech zone