About this document
Amharic Speech Recognition Proposal by Addis Belayhun is a document available to read on EtoBox.
The document proposes developing an Amharic speech recognition system. It discusses collecting and preprocessing a dataset of transcribed Amharic speech audio from the internet and manually. The methodology will use deep learning libraries and pretrained models to train a model on GPUs. The expected benefits include enabling hands-free work, rapid returns on investment, and assisting people with disabilities. The timeline shows tasks over several months. Risks include lack of high quality, large dataset and
- Author
- Addis Belayhun
- Language
- EN