In a race to deliver the best in AI, Meta takes a step ahead to make communication easier for people speaking different languages. The company has unveiled the latest AI translation tool update to make conversations more spontaneous. Its new AI model supports multiple languages in text and speech, which is set to detect and add the expression to the translation. Moreover, it comes close to the real-time translation with a two-second delay for speech translation.
What is SeamlessM4T v2?
SeamlessM4T is a translation model that Meta released in August 2023. It is showcasing Seamless Communication’s potential with the new features in the latest ‘v2’ architecture with the expansion. It currently supports around 100 languages for texting and 36 for speech recognition. This model will work with speech-to-speech, speech-to-text, and text-to-speech translation and deliver results with automatic speech recognition.
“SeamlessExpressive” is one of the two breakthroughs for Seamless Communication. It attempts to break free from the robotic voice and add expressions to make it more relatable to the user. As it adds the speaking volume, pitch, emotional tone, and pauses to the translated voice so the listener can easily understand the emotion.
Check out this demo page for Seamless Communication for English, Spanish, French, and German language demos.
The second feature of Seamless Communication is “SeamlessStreaming”. It makes sure you hear a translation as soon as the speaker speaks. Now, you don’t have to wait for a sentence or phrase to be completed for its translation. With a slight delay of two seconds, you can quickly hear the translated speech. Meta has clarified that it is difficult to begin translation without delay, as different languages hold distinct structures for sentence formation. Therefore, they have worked on designing a dedicated algorithm to understand whether the audio input has sufficient context to start producing translated output or not.
What can you expect?
Meta’s Seamless Translation AI tool enables real-time communication between different languages. We can expect this to help us communicate efficiently with people speaking different languages. Integrating Smart Glasses can bring revolutionary changes to the AI interpreter model. With the “SeamlessExpression,” more and more people can watch content from various languages and connect globally.
However, Meta has not provided any dates for its public release. Hence, we are unsure when this technology will be available for everyday use. With the new ability to easily connect the user with others through speech translation, hopefully, people will embrace this shift.
What are your thoughts on it? Let us know in the comments below. Follow us on Facebook, Twitter, YouTube, Instagram, Flipboard, and Pinterest.
Subscribe Now & Never Miss The Latest Tech Updates!