How to convert speech to text

Hello,

I am doing a project on speech to text for any one language(except english) without using any APIs or pre-trained model. How can I do that?

Or how can I build my own speech to text model in MATLAB ?

Any kind of help would be appreciated.

Answer 1

[VINAYAK LUHA]

Hi Aviral,

I understand that you want to build your own speech-to-text-model from scratch in MATLAB. If you don't have a model ready for implementation in MATLAB, you may want to refer to some of the existing state-of-the-art speech recognition model and convert their code to MATLAB, Following is a good resource to learn about different speech recognition model - https://paperswithcode.com/task/speech-recognition

Additionally, if you are interested the default model which the MATLAB function "speech2text" uses is the following -Baevski, Alexei, Henry Zhou, Abdelrahman Mohamed, and Michael Auli. “Wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations,” 2020. https://doi.org/10.48550/ARXIV.2006.11477.

Hope this answers your question

Regards,

Vinayak Luha