Road terrain recognition based on tire noise for autonomous vehicle DOI Creative Commons
Dongsheng Yang, Dongmin Zhang,

Yi Yuan

и другие.

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Дек. 28, 2024

Abstract Effective road terrain recognition is crucial for enhancing the driving safety, passability, and comfort of autonomous vehicles. This study addresses challenges accurately identifying diverse surfaces using deep learning in complex environments. We introduce a novel end-to-end Tire Noise Recognition Residual Network (TNResNet) integrated with time-frequency attention module, designed to capture leverage information from tire noise signals classification. Our method was evaluated on five distinct types: asphalt, cement, grass, mud, sand. The performance TNResNet rigorously compared against traditional machine techniques, including Decision Trees, K-Nearest Neighbors, Support Vector Machines, as well advanced models like Long Short-Term Memory Convolutional Neural Networks. Experimental results demonstrate that achieves superior classification accuracy 99.48%, outperforming all comparative methods. work not only establishes robust framework identification but also showcases significant practical implications realm vehicle navigation.

Язык: Английский

Enhancing Cloud Security and Efficiency Through AI-Driven Intrusion Detection and Machine Learning-Based Resource Management DOI
C.P. Ravikumar, Satyanarayana Nimmala, Isha Batra

и другие.

Advances in information security, privacy, and ethics book series, Год журнала: 2025, Номер unknown, С. 239 - 254

Опубликована: Фев. 7, 2025

Cloud computing is essential to modern IT infrastructure but faces challenges in security and resource optimization. This chapter explores enhancing cloud environments using artificial intelligence (AI) machine learning (ML). AI-driven intrusion detection systems (IDS) employ anomaly predictive analytics mitigate threats real time, fortifying against sophisticated attacks. Simultaneously, ML-based management optimizes performance by analyzing usage patterns predicting demands, ensuring cost-efficiency. The highlights methodologies like deep reinforcement learning, illustrating their application improving scalability. Emerging trends such as federated quantum are also discussed, emphasizing the critical role of AI ML advancing sustainable resilient ecosystems.

Язык: Английский

Процитировано

0
Ensuring Driving and Road Safety of Autonomous Vehicles Using a Control Optimiser Interaction Framework Through Smart “Thing” Information Sensing and Actuation DOI Creative Commons
Ahmed Almutairi, Abdullah Faiz Al Asmari, Tariq Alqubaysi

и другие.

Machines, Год журнала: 2024, Номер 12(11), С. 798 - 798

Опубликована: Ноя. 11, 2024

Road safety through point-to-point interaction autonomous vehicles (AVs) assimilate different communication technologies for reliable and persistent information sharing. Vehicle resilience consistency require novel sharing knowledge retaining driving pedestrian safety. This article proposes a control optimiser framework (COIF) organising transmission between the AV interacting “Thing”. The relies on neuro-batch learning algorithm to improve measure’s adaptability with “Things”. In information-sharing process, maximum extraction utilisation are computed track precise environmental knowledge. interactions batched type of traffic obtained, such as population, accidents, objects, hindrances, etc. Throughout travel, vehicle’s rate surrounding environment’s familiarity it classified. neurons connected actuated sensed by identify any unsafe vehicle activity in unknown or unidentified scenarios. Based risk parameters, safe is categorised rate. Therefore, minor changes vehicular decisions monitored, optimised accordingly retain 7.93% navigation assistance 9.76% high intervals.

Язык: Английский

Процитировано

1
Road terrain recognition based on tire noise for autonomous vehicle DOI Creative Commons
Dongsheng Yang, Dongmin Zhang,

Yi Yuan

и другие.

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Дек. 28, 2024

Abstract Effective road terrain recognition is crucial for enhancing the driving safety, passability, and comfort of autonomous vehicles. This study addresses challenges accurately identifying diverse surfaces using deep learning in complex environments. We introduce a novel end-to-end Tire Noise Recognition Residual Network (TNResNet) integrated with time-frequency attention module, designed to capture leverage information from tire noise signals classification. Our method was evaluated on five distinct types: asphalt, cement, grass, mud, sand. The performance TNResNet rigorously compared against traditional machine techniques, including Decision Trees, K-Nearest Neighbors, Support Vector Machines, as well advanced models like Long Short-Term Memory Convolutional Neural Networks. Experimental results demonstrate that achieves superior classification accuracy 99.48%, outperforming all comparative methods. work not only establishes robust framework identification but also showcases significant practical implications realm vehicle navigation.

Язык: Английский

Процитировано

1