Limecraft is the preferred solution for Automatic Speech Recognition (ASR) solution by many audiovisual professionals. To support the widest possible variety of languages and language flavours, and taking into account the specific use case (such as indexing content, text-based editing, or AI subtitling) and operational constraints (cloud-based or locally installed), Limecraft may engage a different ASR engine. This article explains which ASR engines are currently accessible.
Limecraft offers a range of speech-to-text services optimised for different purposes or languages. To obtain the best results for a specific use case or language, we can switch between the different Automatic Speech Recognition (ASR) engines. For example, some ASR engines are rather optimised for speed (low latency) or price, others are optimised for maximum recognition performance. In specific cases we can go as far as training an ASR engine to a specific purpose.
The following speech to text engines are currently supported
Upon Automatic Speech Recognition, Limecraft post-processes the results to optimise the output for the particular use case. For example, turning transcripts in subtitles involves rewriting numerical formats and currencies.
If you have any specific enquiries, please contact us via [email protected].