Corpus word count audio

[JRMeyer]

>>> haqkiemdaim
[July 6, 2019, 11:54pm]

Hi everyone. I got 2 question on TTS dataset/corpus

My question is that:

1. is high frequency word (number of word) in a corpus is better than
low frequency word when it comes to generating an audio for a
particular text?

2. If i have the word 'curious' as the high-frequency word/audio in my
corpus, how can the model generalize that word/audio when it comes
to TTS? slash
Since the word is most likely to have various different intonation
during recording.

Frequency = how many times a certain word occur in our corpus

[This is an archived TTS discussion thread from discourse.mozilla.org/t/corpus-word-count-audio]

[JRMeyer]

>>> erogol
[July 8, 2019, 8:01am]

Yes in general the model works better for words crowded in the
dataset,especially for languages with different pronounciation than
writing. Therefore, it is better to create your dataset with a good
covarage of words and phonemes and it is better to use phonemes for such
languages.

[Archived Post]

[JRMeyer]

>>> josh_meyer
[August 22, 2019, 12:06am]

--- related question:

How important is it to have a Gaussian-like distribution on clip and
sentence lengths in the training corpus?

If I have a corpus with a distribution on text lengths as in the image
below, what would you recommend doing? The histogram below shows numbers
of characters per line of text:

Remove some shorter utterances? Leave as-is?

63.1

As you can see, the dataset has a lot of short sentences (one, two,
three, and four words long).

I'm having an alignment issue during training and inference, in which
the model produces only the first few words of a longer sentence... is
this issue related?

[Archived Post]

[JRMeyer]

>>> erogol
[August 22, 2019, 1:12pm]
I'd say the problem is related. [Maybe you can sub-sam]ple the dataset
biasing towards longer sentences at the later stages of the training.

[Archived Post]

[JRMeyer]

>>> josh_meyer
[August 22, 2019, 6:21pm]

I'll give it a try, thanks!

[Archived Post]