Advancing 3D point cloud understanding through deep transfer learning: A comprehensive survey

Information Fusion, Journal Year: 2024, Volume and Issue: 113, P. 102601 - 102601

Published: July 27, 2024

Language: Английский

---

A Comprehensive Review of Deep Learning: Architectures, Recent Advances, and Applications

Information, Journal Year: 2024, Volume and Issue: 15(12), P. 755 - 755

Published: Nov. 27, 2024

Deep learning (DL) has become a core component of modern artificial intelligence (AI), driving significant advancements across diverse fields by facilitating the analysis complex systems, from protein folding in biology to molecular discovery chemistry and particle interactions physics. However, field deep is constantly evolving, with recent innovations both architectures applications. Therefore, this paper provides comprehensive review DL advances, covering evolution applications foundational models like convolutional neural networks (CNNs) Recurrent Neural Networks (RNNs), as well such transformers, generative adversarial (GANs), capsule networks, graph (GNNs). Additionally, discusses novel training techniques, including self-supervised learning, federated reinforcement which further enhance capabilities models. By synthesizing developments identifying current challenges, insights into state art future directions research, offering valuable guidance for researchers industry experts.

Language: Английский

---

Automatic Speech Recognition: A survey of deep learning techniques and approaches

International Journal of Cognitive Computing in Engineering, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 1, 2025

Language: Английский

---

Artificial Intelligence for Cochlear Implants: Review of Strategies, Challenges, and Perspectives

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 119015 - 119038

Published: Jan. 1, 2024

Automatic speech recognition (ASR) plays a pivotal role in our daily lives, offering utility not only for interacting with machines but also facilitating communication individuals partial or profound hearing impairments. The process involves receiving the signal analog form, followed by various processing algorithms to make it compatible devices of limited capacities, such as cochlear implants (CIs). Unfortunately, these implants, equipped finite number electrodes, often result distortion during synthesis. Despite efforts researchers enhance received quality using state-of-the-art techniques, challenges persist, especially scenarios involving multiple sources speech, environmental noise, and other adverse conditions. advent new artificial intelligence (AI) methods has ushered cutting-edge strategies address limitations difficulties associated traditional techniques dedicated CIs. This review aims comprehensively cover advancements CI-based ASR enhancement, among related aspects. primary objective is provide thorough overview metrics datasets, exploring capabilities AI this biomedical field, summarizing commenting on best results obtained. Additionally, will delve into potential applications suggest future directions bridge existing research gaps domain.

Language: Английский

---

Enhancing IoT Security with CNN and LSTM-Based Intrusion Detection Systems

Published: April 24, 2024

Protecting Internet of things (IoT) devices against cyber attacks is imperative owing to inherent security vulnerabilities. These vulnerabilities can include a spectrum sophisticated that pose significant damage both individuals and organizations. Employing robust measures like intrusion detection systems (IDSs) essential solve these problems protect IoT from such attacks. In this context, our proposed IDS model consists on combination convolutional neural network (CNN) long short-term memory (LSTM) deep learning (DL) models. This fusion facilitates the classification traffic into binary categories, benign malicious activities by leveraging spatial feature extraction capabilities CNN for pattern recognition sequential retention LSTM discerning complex temporal dependencies in achieving enhanced accuracy efficiency. assessing performance model, authors employed new CICIoT2023 dataset training final testing, while further validating model's through conclusive testing phase utilizing CICIDS2017 dataset. Our achieves an rate 98.42%, accompanied minimal loss 0.0275. False positive (FPR) equally important, reaching 9.17% with F1score 98.57%. results demonstrate effectiveness CNN-LSTM fortifying environments potential threats.

Language: Английский

---

Speech Emotion Recognition Using Dual-Stream Representation and Cross-Attention Fusion

Electronics, Journal Year: 2024, Volume and Issue: 13(11), P. 2191 - 2191

Published: June 4, 2024

Speech emotion recognition (SER) aims to recognize human emotions through in-depth analysis of audio signals. However, it remains challenging encode emotional cues and fuse the encoded effectively. In this study, dual-stream representation is developed, both full training fine-tuning different deep networks are employed for encoding patterns. Specifically, a cross-attention fusion (CAF) module designed integrate output recognition. Using encoders (fully text processing network pre-trained large language network), CAF compared other three modules on databases. The SER performance quantified with weighted accuracy (WA), unweighted (UA), F1-score (F1S). experimental results suggest that outperforms leads promising databases (EmoDB: WA, 97.20%; UA, 97.21%; F1S, 0.8804; IEMOCAP: 69.65%; 70.88%; 0.7084; RAVDESS: 81.86%; 82.75.21%; 0.8284). It also found achieves superior than fully network. future improved could be achieved development multi-stream incorporation multi-branch mechanism

Language: Английский

---

Genetic Algorithm-Guided Diverse Sample Selection with Diffusion-Based Generative Memory for Continual Learning of Acoustics

Published: Jan. 1, 2025

Language: Английский

---

A two-stage federated learning method for personalization via selective collaboration

Computer Communications, Journal Year: 2025, Volume and Issue: unknown, P. 108053 - 108053

Published: Jan. 1, 2025

Language: Английский

---

Redesigning Deep Neural Networks: Bridging Game Theory and Statistical Physics

Published: Jan. 1, 2025

Language: Английский

---

Deep neural networks for speech enhancement and speech recognition: A systematic review

Ain Shams Engineering Journal, Journal Year: 2025, Volume and Issue: 16(7), P. 103405 - 103405

Published: May 1, 2025

Language: Английский

---