An open-source neural speech recognition toolkit providing an end-to-end speech recognition pipeline for transcribing raw audio into text.
BasicSR is an open-source neural speech recognition toolkit for researchers and developers. It is built using deep learning techniques to provide an end-to-end speech recognition pipeline. BasicSR takes raw audio as input and outputs transcribed text.
Some key features of BasicSR:
BasicSR aims to advance speech recognition research by providing an open and flexible toolkit. The goal is to reduce time spent on implementation, so researchers can focus more on novel techniques and model architectures to push the state-of-the-art in speech recognition performance.
Here are some alternatives to BasicSR:
Suggest an alternative ❐