Whisper (OpenAI)

Whisper, an open-source automatic speech recognition system, is designed to transcribe and translate speech in multiple languages into English. Trained on 680,000 hours of multilingual and multitask supervised data collected from the web, it is robust to accents, background noise, and technical language. This simple end-to-end approach is implemented as an encoder-decoder Transformer and is capable of performing language identification and phrase-level timestamps. Whisper is designed to be easy to use and have high accuracy, making it an excellent tool for developers to add voice interfaces to more applications. With its ability to translate audio or video to text with language translation, Whisper is a powerful speech-to-text and translation tool that can help users communicate more effectively across language barriers.