• Meraj Ahmed
• Manoj Yadav

Totally different approaches to developing the AI systems for self-driving vehicles exist and the majority of them are terribly advanced and with very high hardware needs. the answer conferred during this paper proposes the machine learning primarily based system to be as easy as possible with terribly low hardware needs. A simple 3 layers deep, absolutely connected neural network was trained to map input pictures from a front facing QVGA camera to steering commands. supported a input image the neural network ought to opt for one amongst the four offered commands (forward, left, right or stop). With minimum of the coaching knowledge (250 images) the system learned to follow the road ahead and keep in its lane. The system mechanically learns necessary road options with solely the steering angle because the input from the human driver. it had been ne'er expressly trained to observe lines on the road. Compared to rather more complicated solutions like express decomposition of the matter, like lane detection and management, and convolutional neural networks just like the end to finish learning planned by the nVidia this method proven to be astonishingly strong and economical. we tend to try and prove that this approach would result in higher performance and lower hardware necessities so creating the event of the self-driving vehicles easier and more cost effective. easy artificial neural network, just like the one conferred during this paper, is enough for comparatively complicated method like lane keeping.

Artificial intelligence;neural network; self-driving; autonomous vehicle; object detection; laptop vision; haarcascade; robot.

"A Report On Self Driving Car Using Machine Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 6, page no.62-65, June-2019, Available :http://www.jetir.org/papers/JETIR1906298.pdf

"A Report On Self Driving Car Using Machine Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 6, page no. pp62-65, June-2019, Available at : http://www.jetir.org/papers/JETIR1906298.pdf