Gold Rush Optimization-Based Pyramid Vision Transformer with Deep Hybrid Kernel Extreme Learning Machine for Blind Spot Detection in Autonomous Vehicles

View at publisher Summarise with VIRAT
Authors: C. Kavitha, P. Boobalan
Cited by: 0
References: 15

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The increasing complexity of autonomous driving demands precise vision systems to detect blind spots under varying environmental conditions. Existing sensor-based approaches struggle with occlusion, illumination changes, and dynamic objects, limiting perception accuracy. This paper proposes a Gold Rush Optimization-based Pyramid Vision Transformer with Deep Hybrid Kernel Extreme Learning Machine (GRO-PVT-DHKELM) method for enhanced Blind Spot Detection (BSD) in Autonomous Vehicles (AVs). PVT extracts multi-scale spatial features, while the DHKELM performs efficient classification using adaptive kernel fusion. The GRO approach optimally tunes model parameters, improving convergence and detection performance. The objective is to develop a robust, real-time, and adaptive vision-based framework that enhances perception reliability in complex traffic scenarios. Experimental results demonstrate that the GRO-PVT-DHKELM model significantly improves detection accuracy, minimizes false alarms, and strengthens safety awareness, thereby advancing the perceptual intelligence of next-generation AVs.

Fields of Study & Subjects

Autonomous Vehicle Technology and Safety Advanced Neural Network Applications Machine Learning and ELM

Publication details

Journal / source
International Conference on Computing, Communication, Control and Cyber-Physical Systems (I5CPS)
Pages
1–6
Publication date
26 February 2026
Language
EN