Line 1: Line 1:
Due to a kiwi server fail, my pre-lecture notes are not as substantial as I would have liked
+
==Post Lecture Speech notes==
See my post-lecture notes for a more detailed description.
+
*The server failed??? When?? Zach do you know anything about this? --[[User:Mboutin|Mboutin]] 19:45, 3 November 2009 (UTC)
+
*It was from around 2pm till about 5:30pm Tuesday.  When I tried to preview my page that I had started writing, it said something like "Server not available."  --[[User:Pclay|Pclay]]
+
*We will look into this. Thanks for the detailed info Peter! --[[User:Mboutin|Mboutin]] 13:33, 4 November 2009 (UTC)
+
  
 +
= Basic Idea =
 +
* Speech is an acoustic signal, which we approximate as an analog signal.  It is our goal to change this analog signal into a digital so that we can perform various forms of processing on it.
  
Notes for speech lecture:
+
= Parts of Speech =
 +
* Before we jump into the mathematical "deep end" we first need to know the basic building blocks of speech
  
Structure:
+
** A sentence that we hear is made up of syllables (sound) and separations (no sound).  Simply put, a '''syllable''' is a single, uninterrupted sound that forms the rhythmical foundation of a language. For example, the word 'water' has two syllables, 'wa' and 'ter' separated by a tiny break in speech.
-> Basic speech stuff (pipes, fricatives)
+
-> Voiced vs. Unvoiced
+
  1) avg power
+
2) zero crossing
+
-> x(t) -> v(t) => s(t) = conv( x(t), v(t) )
+
  periodic filter  phoneme
+
  pulse
+
  train
+
-> Model vocal tract as a series of tubes
+
  - Going through tube delays the signal (show function)
+
- between tubes (show function)
+
+ This model leads to a transfer function
+
-> Transfer function V(d)
+
  
  Since the vocal tract is a cavity that resonates, it amplifies certain frequencies
+
** If we go down further, each syllable is formed of phonemes. A '''phoneme''' is the smallest, segmental unit of sound. It is what forms the difference between utterances.  Even though two different groups use the same language and have different accents and because the phonemes have the same function.
  X(f) is sum(a_k * delta(f-kf_a))
+
  
This frequencies, which are the local maxes of |S(f)| are called formants
+
Since phonemes are the smallest block of a speech signal, it is no surprise that form the basis for speech analysis.
  
  - Generally, the vocal tract transfer function is an all-pole filter
+
prelecture notes here
    where a real pole or a complex pole pair correspond to a resonance.
+
[[SupplementarySpeech_prelecture]]
  - Also, if you are given a z-model, F = theta / (2*pi*T) where T is
+
    the sampling period. (same thing as wT = theta
+
 
+
  - zeros, anti-resonances, of the transfer function will occur when there is no
+
    measurable output (i.e. Nasals and Fricatives) 
+
  - Nasal => output from the mouth is zero
+
    Fricatives/stop consonants => blockage behind source is infinite (forcing air
+
    through constriction)
+
   
+
-> Spectrograms
+
  - Models frequency vs. time
+
  - Use a short-time DTFT to obtain useful info about an utterance
+
    X_m(e^jw) = sum( x(n)w(n-m)e^(-jwn))
+
  - wideband uses window length = one period
+
    - high time resolution, low freq
+
    - striations due to energy variation
+
  - narrowband captures several periods
+
    - high freq, low time
+
    - striations correspond to peaks in frequency spectrum.
+
 
+
  The formants correspond to the dark bands.
+
 
+
-> How to read a spectrogram by Rob Hagiwara
+
  http://home.cc.umanitoba.ca/~robh/howto.html
+

Revision as of 17:44, 14 November 2009

Post Lecture Speech notes

Basic Idea

  • Speech is an acoustic signal, which we approximate as an analog signal. It is our goal to change this analog signal into a digital so that we can perform various forms of processing on it.

Parts of Speech

  • Before we jump into the mathematical "deep end" we first need to know the basic building blocks of speech
    • A sentence that we hear is made up of syllables (sound) and separations (no sound). Simply put, a syllable is a single, uninterrupted sound that forms the rhythmical foundation of a language. For example, the word 'water' has two syllables, 'wa' and 'ter' separated by a tiny break in speech.
    • If we go down further, each syllable is formed of phonemes. A phoneme is the smallest, segmental unit of sound. It is what forms the difference between utterances. Even though two different groups use the same language and have different accents and because the phonemes have the same function.

Since phonemes are the smallest block of a speech signal, it is no surprise that form the basis for speech analysis.

prelecture notes here SupplementarySpeech_prelecture

Alumni Liaison

has a message for current ECE438 students.

Sean Hu, ECE PhD 2009