Robust Persian Digit Recognition in Noisy Environments Using Hybrid CNN-BiGRU Model

TL;DR

This paper presents a hybrid CNN-BiGRU model for recognizing Persian digits in noisy environments, achieving high accuracy and outperforming existing methods by significant margins.

Contribution

It introduces a novel hybrid CNN-BiGRU approach using word units for noise-robust Persian digit recognition, outperforming phoneme-based models.

Findings

Achieved over 98% accuracy on clean data

Improved recognition by 26.88% in noisy conditions

Outperformed phoneme-based LSTM and MTDRCC+MLP models

Abstract

Artificial intelligence (AI) has significantly advanced speech recognition applications. However, many existing neural network-based methods struggle with noise, reducing accuracy in real-world environments. This study addresses isolated spoken Persian digit recognition (zero to nine) under noisy conditions, particularly for phonetically similar numbers. A hybrid model combining residual convolutional neural networks and bidirectional gated recurrent units (BiGRU) is proposed, utilizing word units instead of phoneme units for speaker-independent recognition. The FARSDIGIT1 dataset, augmented with various approaches, is processed using Mel-Frequency Cepstral Coefficients (MFCC) for feature extraction. Experimental results demonstrate the model's effectiveness, achieving 98.53%, 96.10%, and 95.92% accuracy on training, validation, and test sets, respectively. In noisy conditions, the proposed approach improves recognition by 26.88% over phoneme unit-based LSTM models and surpasses the Mel-scale Two Dimension Root Cepstrum Coefficients (MTDRCC) feature extraction technique along with MLP model (MTDRCC+MLP) by 7.61%.