Forum Replies Created
-
AuthorPosts
-
I have one more question. In slide 55 of the slide pack, it says that once we have the sequence of feature vectors for an unknown word, each model generates this sequence. What I don’t understand is, if the unknown word is “car”, how can a model for “cat” generate the sequence for “car”?
So I understand that the language model is just all the HMMs connected together.
So going back to my example:
------HMM for cat --------- start __/ \___ end \ / ------HMM for bat ---------
I put a token at start and send a token along each model.
For each model, is only one observation generated? Or does another token passing process occur in the model to find the most likely observation?Also I did not mean that the file names always had the frequency in it! I just thought it was a weird task since the frequencies were in the sine files and it was simple to calculate the frequencies.
Can you tell me whether or not this is correct for the 100Hz sine wave:
Its fundamental frequency is 100Hz because, since it’s time between two peaks is 0.01 seconds, there must be 100 cycles per second.
In chapter 2, page 18 of the Peter Ladefoged reading it says:
“There is, however, one point to be careful about: in order to build up larger variation in air pressure, the particles move farther and more rapidly. But this does not mean the peaks of pressure occur more frequently. As you can see in figure 2.1, although one sound has twice the amplitude of the other, the peaks of pressure in both of them are still occurring at the same rate of one every one-hundredth of a second. One of the two tuning forks may be making larger vibrations than the other, but they are both making the same number of complete vibrations pet second.”
I can’t get my head around why a bigger change in air pressure would not have more cycles. There is an example that follows, that tries to explain this but it doesn’t clear it up for me.
-
AuthorPosts