16 h 6 m, 22 ago 2023 ano - Meta introduces SeamlessM4T, a Multimodal AI Model for Speech and Text Translations

Descrição:

Meta has introduced SeamlessM4T, the first all-in-one multilingual multimodal AI translation and transcription model, capable of performing speech-to-text, speech-to-speech, text-to-speech, and text-to-text translations for up to 100 languages. This innovative model aims to bridge communication gaps across different languages, supporting nearly 100 languages for various translation tasks. In line with Meta's commitment to open science, SeamlessM4T is being publicly released under a research license, along with the metadata of SeamlessAlign, the largest open multimodal translation dataset to date. The development of SeamlessM4T represents a significant advancement towards creating a universal language translator, reducing errors and delays in translation, and enhancing the efficiency and quality of the process. The model builds on previous projects and is part of Meta's ongoing effort to foster global communication through AI-powered technology.

Adicionado na linha do tempo:

Data:

Imagens: