Automatic speech recognition for the Nepali language using CNN, bidirectional LSTM and ResNet

TL;DR

This paper introduces a novel end-to-end deep learning model combining CNN, bidirectional LSTM, and ResNet for Nepali speech recognition, achieving a CER of 17.06% on the test set.

Contribution

The paper presents a new neural network architecture for Nepali ASR that outperforms previous models using CNN, BiLSTM, and ResNet with CTC loss and beam search decoding.

Findings

Achieved 17.06% CER on Nepali speech dataset.

Model outperforms other neural network variations tested.

Provides open-source code for reproducibility.

Abstract

This paper presents an end-to-end deep learning model for Automatic Speech Recognition (ASR) that transcribes Nepali speech to text. The model was trained and tested on the OpenSLR (audio, text) dataset. The majority of the audio dataset have silent gaps at both ends which are clipped during dataset preprocessing for a more uniform mapping of audio frames and their corresponding texts. Mel Frequency Cepstral Coefficients (MFCCs) are used as audio features to feed into the model. The model having Bidirectional LSTM paired with ResNet and one-dimensional CNN produces the best results for this dataset out of all the models (neural networks with variations of LSTM, GRU, CNN, and ResNet) that have been trained so far. This novel model uses Connectionist Temporal Classification (CTC) function for loss calculation during training and CTC beam search decoding for predicting characters as the most likely sequence of Nepali text. On the test dataset, the character error rate (CER) of 17.06 percent has been achieved. The source code is available at: https://github.com/manishdhakal/ASR-Nepali-using-CNN-BiLSTM-ResNet.