Parallel Network Speech Emotion Recognition Based on Hybrid Attention Mechanism

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: April 29, 2025

In speech emotion recognition, insufficient feature extraction and single-feature limitations often lead to low recognition accuracy. To address these issues, thesis proposes a parallel network structure with hybrid attention mechanism, integrating multi-scale temporal modeling enhance performance. The model maps 81-dimensional combined features 128 dimensions via an embedding layer, enriching representation for subsequent layers. These are then processed by three networks, each comprising dilated convolution module, bidirectional long short-term memory mechanism. extracts global contextual information, improving long-term dependency capture, while the models dependencies, capturing emotional variations over time. mechanism further refines weighting across channel dimensions. Experiments on RAVDESS dataset demonstrate that proposed method achieves 96.61% accuracy 96.52% precision in 8-class classification task, outperforming traditional convolutional neural network, other attention-based models. results highlight its effectiveness extracting features, accuracy, offering novel solution recognition.

Language: Английский

---

Multi-source variational mode transfer learning for enhanced PM2.5 concentration forecasting at data-limited monitoring stations

Expert Systems with Applications, Journal Year: 2023, Volume and Issue: 238, P. 121714 - 121714

Published: Sept. 27, 2023

Language: Английский

---

Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care for Chronic Diseases: A Systematic Literature Review

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 122557 - 122587

Published: Jan. 1, 2024

Language: Английский

---

Physics‐informed anomaly and fault detection for wind energy systems using deep CNN and adaptive elite PSO‐XGBoost

IET Generation Transmission & Distribution, Journal Year: 2025, Volume and Issue: 19(1)

Published: Jan. 1, 2025

Abstract Wind energy systems require fault diagnosis that identifies faults despite data inconsistencies. This study addresses challenges in supervisory control and acquisition (SCADA) for monitoring wind turbine conditions from imbalanced representation error vulnerability. It examines the efficacy of adaptive elite‐particle swarm optimization (AEPSO)‐tuned extreme gradient boosting (XGBoost) on an SCADA dataset classification. The methodology integrates resampled with t ‐distributed stochastic neighbour embedding represented deep learning features. Employing AEPSO‐XGBoost classifier trained merged a physics‐informed convolutional neural network forms basis (alarm) classification model. regressor is validated across three distinct rear bearing temperature datasets, facilitating parameter model robustness. Also, this explores supervised unsupervised anomaly detection models using PDCNN rear‐bearing data. Findings exhibit substantial prediction enhancements by merging Moreover, results show applying can significantly improve metrics. Through AEPSO‐XGBoost's enhancing within proposes integrated framework as innovative predictive maintenance system systems.

Language: Английский

---

A Novel Hybrid Approach Combining PDEM and Bayesian Optimization Deep Learning for Stochastic Vibration Analysis in Train-Track-Bridge Coupled System

Reliability Engineering & System Safety, Journal Year: 2025, Volume and Issue: unknown, P. 110827 - 110827

Published: Jan. 1, 2025

Language: Английский

---