Self-Driving Car Based CNN Deep Learning Model

Ibrahim, Amin S.; Ibrahim, Omer M.; Na&#039;eem, Ahmed M.; Hanna, Andrew S.; Ezz, Shady M.; Mohamed, Mohamed R.; Abdel Hameed, Hassan M.; Sanad, Felopateer; Hassan, Hossam I.; Refky, Adel; Abdel Ghany, A. M.

doi:10.21608/iugrc.2022.303140

Self-Driving Car Based CNN Deep Learning Model

Document Type : Original Article

Authors

Thebes Higher Institute for Engineering, Egypt.

10.21608/iugrc.2022.303140

Abstract

Artificial Intelligent (AI) technology is capable of thinking and recognizing environmental things based on the vast
amount of historical training data. AI technology can mimic the human brain with large and complicated computations and short processing time. One of the main challenges in our daily life is the rapid growth of accidents and deaths due to the wrong driving by citizens and unrespecting the traffic rules. Thus, AI technology is coming as a solution to solve this issue through the so-called self-driving car. In this paper, we proposed a self-driving car based on the Convolutional Neural Network (CNN) deep learning model in AI technology. The paper designed and implemented a self-driving car prototype to prove the concept and validate our experimental self-driving car model. It is remarked that a porotype is successfully trained and tested about 5000 images with very low training and validation loss of less than 0.05 and 0.12 respectively.

Keywords