The Effects of Image Augmentation on Efficiency of a Convolutional Neural Network of a Self-Driving Car

Authors

  • Kiaan Upamanyu Greenwood High International School Bangalore
  • Indukala P.R. Greenwood High International School Bangalore

DOI:

https://doi.org/10.47611/jsrhs.v10i2.1486

Keywords:

Computer Vision, Artificial Intelligence, Machine Learning, Autonomous Vehicles, Convolutional Neural Networks, Self-driving car, Computer Science

Abstract

The objective of this paper was to study the effectiveness of image augmentation techniques in training a Convolutional Neural Network (CNN) of a self-driving car and identify the most suitable form of image augmentation technique, using the Udacity Car Simulator. Firstly, a dataset of augmented and non-augmented images from a training track, consisting of left-, right-, and front-facing views from the car cameras was created. Various image augmentation techniques were used: zoom, brightness, pan, flip, random (augments the image by arbitrarily choosing a technique from the previous four), and no augmentation. Secondly, training datasets consisting of the aforementioned images and a log of car turning angles, throttle, and brake were built. The final training datasets were then used with NVIDIA training method to train different CNN. The different trained networks generated steering commands from the front-facing camera of the simulation and test track had no effect on the generalization of the CNN. Lastly, different trained networks were used on the test track of Udacity Car Simulator to calculate the following variables: distance travelled, and number of crashes made by the car. After these values were acquired, an efficiency analysis was performed. The results suggested augmentation of training data is a crucial factor when it comes to the process of generalizing a model to perform tasks. Random augmentations performed the best, but a combination of flip and brightness augmentations performed equally efficiently.

Downloads

Download data is not yet available.

Author Biographies

Kiaan Upamanyu, Greenwood High International School Bangalore

Student

Indukala P.R. , Greenwood High International School Bangalore

Computer Science Head of Department 

References or Bibliography

Bojarski, M., Testa, D. D., Dworakowski, D., Firner, B., Flepp, B., Goyal, P., . . . Zieba, K. (2016). End to End Learning for Self-Driving Cars. Nvidia. Retrieved from https://arxiv.org/abs/1604.07316

Brownlee, J. (2019, February 27). Machine Learning Mastery. Retrieved from How to use Learning Curves to Diagnose Machine Learning Model Performance: https://machinelearningmastery.com/learning-curves-for-diagnosing-machine-learning-model-performance/

Canuma, P. (2018, October 11). Towards Data Science. Retrieved from Image Pre-processing: https://towardsdatascience.com/image-pre-processing-c1aec0be3edf

Keras. (n.d.). Layer activation functions. Retrieved from Keras: https://keras.io/api/layers/activations/

Slim , R., Sharaf, A., Slim, J., & Tanveer, S. (2020, September). The Complete Self-Driving Car Course - Applied Deep Learning. The Complete Self-Driving Car Course - Applied Deep Learning. Udemy. Retrieved from https://www.udemy.com/course/applied-deep-learningtm-the-complete-self-driving-car-course/

Slim, R. (2020, May 12). Self-Driving-Car-Course-Codes/Section 14 - Behavioral Cloning/. Retrieved from Github: https://github.com/rslim087a/Self-Driving-Car-Course-Codes/tree/master/Section%2014%20-%20Behavioral%20Cloning

Udacity. (2019, January 11). udacity/self-driving-car-sim. Retrieved from Github: https://github.com/udacity/self-driving-car-sim

Union of Concerned Scientists. (2017, January 26). Union of Concerned Scientists. Retrieved from Self-Driving Cars Explained: https://www.ucsusa.org/resources/self-driving-cars-101

Upamanyu, K. (2020, December 27). Kiaan1204 / TRAINING-DATA-1-AND-2. Retrieved from Github: https://github.com/Kiaan1204/TRAINING-DATA-1-AND-2

Wang, J., & Perez, L. (2017, December 13). The Effectiveness of Data Augmentation in Image Classification using Deep Learning. Retrieved from arXiv: https://arxiv.org/abs/1712.04621

Zhang, A., Lipton, Z. C., Li, M., & Smola, A. J. (2020). Dive into Deep Learning. Retrieved from https://d2l.ai/chapter_computer-vision/image-augmentation.html

Published

07-01-2021

How to Cite

Upamanyu, K., & Ramaswamy, I. P. . (2021). The Effects of Image Augmentation on Efficiency of a Convolutional Neural Network of a Self-Driving Car. Journal of Student Research, 10(2). https://doi.org/10.47611/jsrhs.v10i2.1486

Issue

Section

HS Research Articles