Authors:-Kumtale Aishwarya Bhausaheb , Rutuja Ramesh Madnaik , Navale Shruti Sudhir ,Dr. Muthusundar S.K.
Abstarct:-Most of the car accidents happen because of human errors while driving. Therefore, we think this is the serious real life problem and this needs to be solved. In this project, we focused on designing and implementing an efficient model using convolutional neural network for self-driving car module. Our project entirely focuses on predicting most accurate rotation angle for the cars based on the input image feed. The calculation supports both types of images taken on highway or on local roads. We implement the models for predicting steering angles on the Udacity dataset and visualize the results by using the dataset given by comma ai. For better performance, our system can run on the GPU using CUDA to process more and more frames to give accurate results quickly. CUDA enables increase in computing performance by harnessing power of GPU.
Keywords:-Machine Learning, Convolutional Neural Network, Steering Angle Prediction
Full Issue
| View or download the full issue | PDF 38-42 |