Deployment of Artificial Intelligence (AI) in daily life will show a new way to the future, the complex task will become easy. One of such evolution of AI technology is the creation of the Self-Driving cars. But according to the statistics, there are some cases where the Self-Driving cars killed people due to the failure in the prediction or hardware problem. So, to overcome those kinds of situations the AI models should be accurate and also humans should take part in making them perfect. For example, when there are some passengers in a self-driving car and the car has lost its control due to the fault in the hardware of the computer installed in the car then an authorized person in the car should be able to instruct it using voice commands to control the car at that moment. The main focus of this paper relies on the development of Self-Driving Cars using the Convolutional Neural Networks (CNN), drawbacks and solutions for the drawbacks. The Self-Driving Car CNN model was trained using the Asphalt-8 game data and the Voice command prediction CNN model was trained with 3 different persons (1-Kid, 1-Man, 1-Woman) voices. The accuracy of both the CNN models were 99% and was tested on the same game where they have produced their best results.

No abstract

No abstract

No abstract

No abstract