In the narration algorithm, can a few commonly used words be trained/fixed? Like “EA” comes out as “eeeyahhh”.
Yes, basically just adjust the lexicon or use an alias as described in the AWS backend being used.
Or do a regex (replace “EA”) with a phonetic transcription SSML tag, described here:
https://aws.amazon.com/blogs/machine-learning/customize-pronunciation-using-lexicons-in-amazon-polly/
In the narration algorithm, can a few commonly used words be trained/fixed? Like “EA” comes out as “eeeyahhh”.
Yes, basically just adjust the lexicon or use an alias as described in the AWS backend being used.
Or do a regex (replace “EA”) with a phonetic transcription SSML tag, described here:
https://aws.amazon.com/blogs/machine-learning/customize-pronunciation-using-lexicons-in-amazon-polly/