Artificial Intelligence is booming
exponentially and talk of the
town technology since recent years. Researchers and scientists truly believe
that AI would be the future technology that dominates all over the globe in all
aspects of fields.
As a next step towards AI technology, Google has developed a text-to-speech
system like a human speech articulation.
The text-to-speech is called as “Tacotron 2”, that delivers an AI
generated computer speech that almost matches with the voice of humans.
Google's CEO Mr. Sundar Pichai announced at the Google i/o developers conference that the company is shifting its focus
from mobile-first to AI-first . And it has launched several products and
features based on AI technology like Google Lens, Smart Reply for Gmail and
Google Assistant for iPhone.
The system first creates a spectrogram
of the text, a visual representation of how the speech should sound – according
to the paper published.
That image is put through Google's
existing WaveNet algorithm, which uses the image and brings AI closer than ever
to clearly distinguish imitate or mimicking human speech. The algorithm can easily
learn different voices and even generates artificial breaths.
The researchers quoted that the MOS – Mean Opinion Score for the model
is 4.53, comparable to a MOS of 4.58.
"Tacotron 2" can detect from
context the difference between the noun "desert" and the verb
"desert," as well as the noun "present" and the verb
"present," and alter its pronunciation accordingly.
It can place emphasis on capitalized
words and apply the proper inflection when asking a question rather than making
a statement.
Still the developers of Google did not reveal much information to figure
out how far they have come in developing this system.
According to the paper, “Based on the paper, it's highly probable that
"gen" indicates speech generated by Tacotron 2 and "gt" is
real human speech. ("GT" likely stands for "ground truth,"
a machine learning term that basically means "the real deal".”
No comments:
Post a Comment