I'm curious about "frame blocking" which is the first step of feauture extraction as i saw in web site. It says that these steps should be follow for feature extraction of an audio:
What does "frame blocking" mean? Can anyone give an algorithm or code(c#) for that?
By the way, is this order of step true for feature extraction?
What should I do as "pre-filtering" an aduio before feature extraction?
Thanks.
Here is what I found from a google search on the term, "frame blocking" - which google auto-completed to "frame blocking of speech signal".
Relevant text:
The most common approaches in speach signal processing are based on short-time analysis. The preemphased signal is blocked into frames of N samples. Frame duration typically ranges between 10 - 30 mxec. Values in this range represent a tradeoff between the rate of change of spectrum and system complexity. The proper frame duration is ultimately dependent on the velocity of the articulators in the speech production syste. Some sounds (e.g., stop consonants) exhibit sharp spectral transition whcih can result in spectral peaks shifting as much as 80 Hz/msec (Mar76]. Figure 2.3 illustrates the blocking of a word in to J frames. The amount of overlap to some extent controls how quickly parameters can change from frame to frame.
Now the question to you is, what have you tried with regard to frame blocking?
