Bangla Speech Recognition

This project implements a Sequence-to-Sequence (Seq2Seq) model for Bangla speech recognition using PyTorch. The project focuses on sequence-to-sequence tasks with utilities for data loading, preprocessing, training, and validation. This project aims to build an efficient speech recognition system for the Bangla language.

Key Features:

• Seq2Seq Model: A sequence-to-sequence deep learning model is used to transcribe Bangla speech to text.
• PyTorch Implementation: The entire model is built using the PyTorch framework.
• Data Loading & Preprocessing: The project includes utilities for loading, preprocessing, and preparing the speech dataset.
• Training and Validation: Scripts are provided for training and validating the model.

Technologies and Dependencies:

• Python 3.7
• PyTorch
• NumPy
• Librosa: For audio processing.

You can view the full project and source code on GitHub: Bangla Speech Recognition