Unit Selection

From SpeechWiki

(Difference between revisions)
Jump to: navigation, search
(New page: ==Error driven unit selection== ===Some statistics=== {| class="wikitable" |+Size of phoneme corpus ! Utterances | 2000 |- ! Phonemes (including EOW and <s> </s> words) | 95938 |- ! Fra...)
Line 17: Line 17:
[[Image:PhoneConfusionMatrix.png|thumb|none|800px|The Confusion matrix]]
[[Image:PhoneConfusionMatrix.png|thumb|none|800px|The Confusion matrix]]
[[Image:PhoneConfusionMatrixTop40.png|thumb|none|800px| The confusion matrix ignoring EOW and various non-speech events]]
[[Image:PhoneConfusionMatrixTop40.png|thumb|none|800px| The confusion matrix ignoring EOW and various non-speech events]]
 +
 +
The number of triphone decision tree leaf nodes is essentially not correlated with any of {total Frames,total Phones, error Phones, error Frames} per phone.  Since the units are chosen to maximally cover the mistakes in the corpus, this might suggest that deepening the triphone DTs is not the same thing as coming up with these error-based units.
==Syllable Units==
==Syllable Units==
[[Category:Fisher Experiments]]
[[Category:Fisher Experiments]]

Revision as of 02:06, 8 February 2009

Error driven unit selection

Some statistics

Size of phoneme corpus
Utterances 2000
Phonemes (including EOW and words) 95938
Frames 756009
The probability of a particular phone occurring in the corpus, and the probability of a particular phone that is correct in the corpus
The Confusion matrix
The confusion matrix ignoring EOW and various non-speech events

The number of triphone decision tree leaf nodes is essentially not correlated with any of {total Frames,total Phones, error Phones, error Frames} per phone. Since the units are chosen to maximally cover the mistakes in the corpus, this might suggest that deepening the triphone DTs is not the same thing as coming up with these error-based units.

Syllable Units

Personal tools