Electronic Thesis/Dissertation
 

Dual Vision: Enhancing Autonomous Navigation with AutoGuardian CNN and YOLO Object Detection Synergy

Open Access Deposited

In the growing field of autonomous vehicle technology, the dependability and precision of computer vision systems are crucial. This study compares the performance of two advanced object detection models: the bespoke AutoGuardian Convolutional Neural Network (CNN) and the popular YOLOv8 model. The main goal is to explore the AutoGuardian CNN model's potential as a verification layer to improve the YOLO Object Detection model, thus boosting the effectiveness of computer vision in autonomous vehicles. The study uses a thorough evaluation framework with standard measures such as accuracy, precision, and recall, and it assesses the influence of the AutoGuardian CNN as a verification layer on the YOLO model's output.Initial results indicate that including the AutoGuardian CNN as a verification layer shows great potential in reducing incorrect identifications and improving the overall dependability of the YOLOv8 model in object detection across different environmental settings. This study enhances the scholarly discussion on object detection models in self- driving cars and provides valuable guidance for creating and implementing more robust and dependable computer vision systems in safety-critical scenarios. The study seeks to explore the complex relationship between various object identification models and their ability to operate in the presence of visual noise, with the goal of advancing autonomous vehicle technology. This research has consequences that go beyond autonomous driving, providing vital insights for the wider field of computer vision and its use in safety-critical systems.

Author Language Keyword Date created Type of Work License
  • All rights reserved
Rights statement GW Unit Degree Advisor Committee Member(s) Persistent URL

Notice to Authors

If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.

Thumbnail Title Date Uploaded Visibility Actions
Preview of Boadu_gwu_0075A_16800.pdf Boadu_gwu_0075A_16800.pdf 2024-10-02 Open Access