Ameliorated End-to-End Deep Learning System for Self Learning Cars

R, Thirumahal and T, BalajiMuthazhagan (2021) Ameliorated End-to-End Deep Learning System for Self Learning Cars. In: ICCAP 2021, 7-8 December 2021, Chennai, India.

[thumbnail of PDF]
Preview
Text (PDF)
eai.7-12-2021.2314534.pdf - Published Version

Download (668kB) | Preview

Abstract

The introduction of self-driving automobiles in today's society necessitates the development of high-quality algorithms to guide them. Convolutional neural networks are extensively utilised because of their ability to classify images based on observed attributes. The proposed solution entails steering a car autonomously using just the windshield view as input. This is accomplished by using Convolutional neural networks in an end-to-end deep learning strategy, as proposed by NVIDIA for self-driving automobiles. To improve the accuracy of existing models, the neural network will be supported with inputs from image processing modules that identify lane markers and cars.Reduced costs, increased safety, and increased mobility are all advantages of providing such an algorithm for retrofitting existing cars.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: autonomous vehicles self-driving car end-to-end deep learning convolutional neural networks feature extraction classification tools
Subjects: T Technology > T Technology (General)
Depositing User: EAI Editor IV
Date Deposited: 12 Jan 2022 11:48
Last Modified: 12 Jan 2022 11:48
URI: https://eprints.eudl.eu/id/eprint/9887

Actions (login required)

View Item
View Item