Cross corpus multi-lingual speech emotion recognition using ensemble learning

Complex & Intelligent Systems, Journal Year: 2021, Volume and Issue: 7(4), P. 1845 - 1854

Published: Jan. 11, 2021

Abstract Receiving an accurate emotional response from robots has been a challenging task for researchers the past few years. With advancements in technology, like service interact with users of different cultural and lingual backgrounds. The traditional approach towards speech emotion recognition cannot be utilized to enable robot give efficient response. conventional uses same corpus both training testing classifiers detect emotions, but this generalized multi-lingual environments, which is requirement used by people all across globe. In paper, series experiments are conducted highlight ensemble learning effect using majority voting technique cross-corpus, system. A comparison performance against machine algorithms performed. This study tests classifier’s trained on one data another evaluate its efficiency detection. According experimental analysis, highest accuracy corpora. Using gives benefit combining classifiers’ instead choosing classifier compromising certain language corpus’s accuracy. Experiments show increased 13% Urdu corpus, 8% German 11% Italian 5% English with-in testing. For cross-corpus experiments, improvement 2% when 15% achieved. An increase 7% obtained data, 3% data. prove that promising results other state-of-the-art techniques.

Language: Английский

---

Assessing and controlling infection risk with Wells-Riley model and spatial flow impact factor (SFIF)

Sustainable Cities and Society, Journal Year: 2021, Volume and Issue: 67, P. 102719 - 102719

Published: Jan. 18, 2021

Language: Английский

---

COVID-19 detection using federated machine learning

PLoS ONE, Journal Year: 2021, Volume and Issue: 16(6), P. e0252573 - e0252573

Published: June 8, 2021

The current COVID-19 pandemic threatens human life, health, and productivity. AI plays an essential role in case classification as we can apply machine learning models on data to predict infectious cases recovery rates using chest x-ray. Accessing patient’s private violates patient privacy traditional model requires accessing or transferring whole train the model. In recent years, there has been increasing interest federated learning, it provides effective solution for privacy, centralized computation, high computation power. this paper, studied efficacy of versus by developing two (a a model)using Keras TensorFlow federated, used descriptive dataset x-ray (CXR) images from patients. During training stage, tried identify which factors affect prediction accuracy loss like activation function, optimizer, rate, number rounds, Size, kept recording plotting per each round, performance, found that softmax function SGD optimizer give better loss, changing rounds rate slightly effect but size did not have any loss. finally, build comparison between proposed models’ accuracy, performance speed, results demonstrate higher time than

Language: Английский

---

Application of Deep Learning Techniques in Diagnosis of Covid-19 (Coronavirus): A Systematic Review

Neural Processing Letters, Journal Year: 2022, Volume and Issue: 55(3), P. 3551 - 3603

Published: Sept. 16, 2022

Language: Английский

---

A Novel Framework Based on Deep Learning and ANOVA Feature Selection Method for Diagnosis of COVID-19 Cases from Chest X-Ray Images

Computational Intelligence and Neuroscience, Journal Year: 2022, Volume and Issue: 2022, P. 1 - 11

Published: Jan. 7, 2022

Background and Objective. The new coronavirus disease (known as COVID-19) was first identified in Wuhan quickly spread worldwide, wreaking havoc on the economy people’s everyday lives. As number of COVID-19 cases is rapidly increasing, a reliable detection technique needed to identify affected individuals care for them early stages reduce virus’s transmission. most accessible method identification Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR); however, it time-consuming has false-negative results. These limitations encouraged us propose novel framework based deep learning that can aid radiologists diagnosing from chest X-ray images. Methods. In this paper, pretrained network, DenseNet169, employed extract features Features were chosen by feature selection method, i.e., analysis variance (ANOVA), computations time complexity while overcoming curse dimensionality improve accuracy. Finally, selected classified eXtreme Gradient Boosting (XGBoost). ChestX-ray8 dataset train evaluate proposed method. Results Conclusion. reached 98.72% accuracy two-class classification (COVID-19, No-findings) 92% multiclass No-findings, Pneumonia). method’s precision, recall, specificity rates 99.21%, 93.33%, 100%, respectively. Also, achieved 94.07% 88.46% 100% classification. experimental results show outperforms other methods be helpful diagnosis cases.

Language: Английский

---

Challenges and strategies for wide-scale artificial intelligence (AI) deployment in healthcare practices: A perspective for healthcare organizations

Artificial Intelligence in Medicine, Journal Year: 2024, Volume and Issue: 151, P. 102861 - 102861

Published: March 30, 2024

Language: Английский

---

Review on the Evaluation and Development of Artificial Intelligence for COVID-19 Containment

Sensors, Journal Year: 2023, Volume and Issue: 23(1), P. 527 - 527

Published: Jan. 3, 2023

Artificial intelligence has significantly enhanced the research paradigm and spectrum with a substantiated promise of continuous applicability in real world domain. intelligence, driving force current technological revolution, been used many frontiers, including education, security, gaming, finance, robotics, autonomous systems, entertainment, most importantly healthcare sector. With rise COVID-19 pandemic, several prediction detection methods using artificial have employed to understand, forecast, handle, curtail ensuing threats. In this study, recent related publications, methodologies medical reports were investigated purpose studying intelligence's role pandemic. This study presents comprehensive review specific attention machine learning, deep image processing, object detection, segmentation, few-shot learning studies that utilized tasks COVID-19. particular, genetic analysis, clinical data sound biomedical classification, socio-demographic anomaly health monitoring, personal protective equipment (PPE) observation, social control, patients' mortality risk approaches forecast threatening factors demonstrates artificial-intelligence-based algorithms integrated into Internet Things wearable devices quite effective efficient forecasting insights which actionable through wide usage. The results produced by prove is promising arena can be applied for disease prognosis, forecasting, drug discovery, development sector on global scale. We indeed played important helping fight against COVID-19, insightful knowledge provided here could extremely beneficial practitioners experts domain implement systems curbing next pandemic or disaster.

Language: Английский

---

A medical multimodal large language model for future pandemics

npj Digital Medicine, Journal Year: 2023, Volume and Issue: 6(1)

Published: Dec. 2, 2023

Deep neural networks have been integrated into the whole clinical decision procedure which can improve efficiency of diagnosis and alleviate heavy workload physicians. Since most are supervised, their performance heavily depends on volume quality available labels. However, few such labels exist for rare diseases (e.g., new pandemics). Here we report a medical multimodal large language model (Med-MLLM) radiograph representation learning, learn broad knowledge image understanding, text semantics, phenotypes) from unlabelled data. As result, when encountering disease, our Med-MLLM be rapidly deployed easily adapted to them with limited Furthermore, supports data across visual modality chest X-ray CT) textual free-text note); therefore, it used tasks that involve both We demonstrate effectiveness by showing how would perform using COVID-19 pandemic "in replay". In retrospective setting, test early datasets; in prospective variant COVID-19-Omicron. The experiments conducted 1) three kinds input data; 2) downstream tasks, including disease reporting, diagnosis, prognosis; 3) five 4) different languages, English, Chinese, Spanish. All show make accurate robust decision-support little labelled

Language: Английский

---

A Survey of Deep Learning Techniques for the Analysis of COVID-19 and their usability for Detecting Omicron

Journal of Experimental & Theoretical Artificial Intelligence, Journal Year: 2023, Volume and Issue: 36(8), P. 1779 - 1821

Published: Jan. 12, 2023

The Coronavirus (COVID-19) outbreak in December 2019 has drastically affected humans worldwide, creating a health crisis that infected millions of lives and devastated the global economy. COVID-19 is ongoing, with emergence many new strains. Deep learning (DL) techniques have proven helpful efficiently analysing delineating infectious regions radiological images. This survey paper draws taxonomy deep for detecting infection radiographic imaging modalities Chest X-Ray, Computer Tomography. DL are broadly categorised into classification, segmentation, multi-stage approaches diagnosis at image region-level analysis. These further classified as pre-trained custom-made Convolutional Neural Network architectures. Furthermore, discussion drawn on datasets, evaluation metrics, commercial platforms provided detection. In end, brief look paid to emerging ideas, gaps existing research, challenges developing diagnostic techniques. provides insight promising areas research likely guide community upcoming development COVID-19. will pave way accelerate designing customised DL-based tools effectively dealing variants challenges.

Language: Английский

---

Deep learning for intelligent demand response and smart grids: A comprehensive survey

Computer Science Review, Journal Year: 2024, Volume and Issue: 51, P. 100617 - 100617

Published: Feb. 1, 2024

Electricity is one of the mandatory commodities for mankind today. To address challenges and issues in transmission electricity through traditional grid, concepts smart grids demand response have been developed. In such systems, a large amount data generated daily from various sources as power generation (e.g., wind turbines), distribution (microgrids fault detectors), load management (smart meters electric appliances). Thanks to recent advancements big computing technologies, Deep Learning (DL) can be leveraged learn patterns predict peak hours. Motivated by advantages deep learning grids, this paper sets provide comprehensive survey on application DL intelligent response. Firstly, we present fundamental DL, response, motivation behind use DL. Secondly, review state-of-the-art applications including forecasting, state estimation, energy theft detection, sharing trading. Furthermore, illustrate practicality via cases projects. Finally, highlight presented existing research works important potential directions

Language: Английский

---