Tools
Tools
Singh, Sumit (2022) A Deep Learning Model for Irish English and Hindi Language Identification. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (1MB) | Preview |
Preview |
PDF (Configuration manual)
Download (1MB) | Preview |
Abstract
Spoken Language Identification is a process of recognizing languages based on audio samples. Language identification models help enable speech-based applications that ease the use of technology for people who find modern technology challenging. This research proposes a deep learning language identification model that can identify and differentiate between three languages Irish, English and Hindi. The proposed deep learning model performs feature extraction based on the frequency and pitch of an audio sample represented by a mel spectrogram using a convolutional neural network (CNN). The audio samples are of varying sizes from 1 second to 10 seconds. The audio sample are used to create the RGB spectrumbased spectrograms. The spectrograms are processed using data augmentation techniques. Pre-trained models such as Resnet50, InceptionV3, EfficientNet-B0 are also trained along with the proposed CNN model and evaluated based on the accuracy, recall, precision and loss values. The CNN model achieves best with an accuracy of 93.50% on the test dataset. This research will help create faster and more efficient speech-based applications in Irish, English and Hindi languages.
| Item Type: | Thesis (Masters) |
|---|---|
| Subjects: | Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning |
| Divisions: | School of Computing > Master of Science in Data Analytics |
| Depositing User: | Tamara Malone |
| Date Deposited: | 13 Mar 2023 10:50 |
| Last Modified: | 13 Mar 2023 10:50 |
| URI: | https://norma.ncirl.ie/id/eprint/6315 |
Actions (login required)
 |
View Item |