Regularized Urdu Speech Recognition with Semi-Supervised Deep Learning

Abstract

Automatic Speech Recognition, (ASR) has achieved the best results for English, with end-to-end neural network based supervised models. These supervised models need huge amounts of labeled speech data for good generalization, which can be quite a challenge to obtain for low-resource languages like Urdu. Most models proposed for Urdu ASR are based on Hidden Markov Models (HMMs). This paper proposes an end-to-end neural network model, for Urdu ASR, regularized with dropout, ensemble averaging and Maxout units. Dropout and ensembles are averaging techniques over multiple neural network models while Maxout are units in a neural network which adapt their activation functions. Due to limited labeled data, Semi Supervised Learning (SSL) techniques are also incorporated to improve model generalization. Speech features are transformed into a lower dimensional manifold using an unsupervised dimensionality-reduction technique called Locally Linear Embedding (LLE). Transformed data along with higher dimensional features is used to train neural networks. The proposed model also utilizes label propagation-based self-training of initially trained models and achieves a Word Error Rate (WER) of 4% less than that reported as the benchmark on the same Urdu corpus using HMM. The decrease in WER after incorporating SSL is more significant with an increased validation data size.

COVID-19 Open Source Data Sets: A Comprehensive Survey

“…Currently, the web repository contains 3D CT images of 10 confirmed COVID-19 cases shared for scientific purposes. The British Society of Thoracic Imaging (BSTI) in collaboration with Cimar UK's Imaging Cloud Technology deployed a free to use encrypted and anonymised online portal to upload and download medical images related to COVID-19 positive and suspected patients 9 . The uploaded images are sent to a group of BSTI experts for diagnosis.…”

COVID-19 open source data sets: a comprehensive survey

“…The SIRM is hosting radiographical images of COVID-19 cases. 9 Their data set has been utilized by some of the cited works in this article. Another open source repository for COVID-19 radiographic images is Radiopaedia.…”

“…Artificial Intelligence (AI) and Machine Learning (ML) techniques have been prominently used to efficiently solve various computer science problems ranging from bio-informatics to image processing. ML is based on the premise that an intelligent machine should be able to learn and adapt from its environment based on its experiences without explicit programming [8,9]. ML models and algorithms have been standardized across multiple programming languages such as, Python and R. The main challenge to the application of ML models is the availability of the open source data [10,11].…”

Prediction of Structural Type for City-Scale Seismic Damage Simulation Based on Machine Learning

“…Subsequently, the designed supervised ML solution indicated that the decision forest model has the best performance for the prediction of structural type; therefore, the decision forest model was selected. Finally, a self-training process [30][31][32] was performed iteratively until the prediction accuracy was accepted. Specifically, the ML model was trained by the sample data, and then the prediction of the trained model was scored and evaluated, separately.…”

“…training process [30][31][32] was performed iteratively until the prediction accuracy was accepted. Specifically, the ML model was trained by the sample data, and then the prediction of the trained model was scored and evaluated, separately.…”