Google Unveils Translatotron, Its Speech-to-Speech Translation System
3:51:18 2019-05-16 996

Google has announced its first direct speech-to-speech translation system called "Translatotron" that can convert verbal communication from one language to another while maintaining the speaker's voice and tempo.

 

Translatotron is based on a sequence-to-sequence network which takes source spectrograms -- a visual representation of frequencies -- as input and generates spectrograms of the translated content in the target language, Ye Jia and Ron Weiss, software engineers at Google Artificial Intelligence (AI) wrote in a blog post on Wednesday.

 

The model makes use of two other separately trained components -- a neural vocoder that converts output spectrograms to time-domain waveforms and a speaker encoder that can be used to maintain the character of the source speaker's voice in the synthesised translated speech.

 

For now, the results of Google's demonstration of the translation system lag behind a conventional cascade system, but the search engine giant said it has "demonstrated the feasibility of the end-to-end direct speech-to-speech translation".

 

 

"By incorporating a speaker encoder network, Translatotron is also able to retain the original speaker's vocal characteristics in the translated speech, which makes the translated speech sound more natural and less jarring," Jia and Weiss said.

 

The company said that Translatotron is the first end-to-end model that can directly translate speech from one language into another.

 

"We hope that this work can serve as a starting point for future research on end-to-end speech-to-speech translation systems," the blog post noted.

 

Lately, Google has been working aggressively on its translation systems.

 

Adding more languages to its real-time translation feature, earlier in 2018, the company introduced accents in Google Translate with a variety of languages in region-based pronunciations.

 

In February, Google extended the "Interpreter" mode on its Assistant to Home speakers and other third-party devices to help users have real-time conversations with compatible devices and translate them into 26 listed languages.

Why Do We Learn?   2025-08-15
Forgive Others   2025-07-23
Reality Of Islam

A Mathematical Approach to the Quran

10:52:33   2024-02-16  

mediation

2:36:46   2023-06-04  

what Allah hates the most

5:1:47   2023-06-01  

allahs fort

11:41:7   2023-05-30  

striving for success

2:35:47   2023-06-04  

Imam Ali Describes the Holy Quran

5:0:38   2023-06-01  

livelihood

11:40:13   2023-05-30  

silence about wisdom

3:36:19   2023-05-29  

MOST VIEWS

Importance of Media

9:3:43   2018-11-05

Illuminations

do not burn out

2:34:48   2022-01-18

the 1st ever brothers

6:14:17   2018-06-21

belief cause cleanliness

10:47:11   2022-11-22

apologize when you are wrong

7:6:7   2022-03-21

your life

2:11:12   2022-10-15

your actions

2:5:14   2023-01-28

your character

2:33:4   2023-02-15



IMmORTAL Words
LATEST There is a Surprisingly Easy Way to Remove Microplastics in Your Drinking Water NASA Probe Could Intercept Interstellar Comet, Scientists Say Ants Are Superefficient at Teamwork, Scientists Discover Improving Quality of Life Interpretation of Sura Maryam (Mary) - Verses 6-8 Let us Not Abuse the Blessings of Allah to Commit Sins This Diet Helps Lower Dementia Risk, And We May Finally Know Why Solar Power Reimagined: New Black Metal Device Generates 15x More Electricity Scientists Uncover Mysterious Oases on the Drying Shores of the Great Salt Lake Why Do We Learn? Interpretation of Sura Maryam - Verses 4-5 A Liar, Not Different from a Dead Body