J. ICT Res. Appl., Vol. 17, No. 2, 2023, 181-200 181
Received August 23
rd
, 2022, 1
st
Revision March 4
th
, 2023, 2
nd
Revision June 15
th
, 2023, Accepted for
publication July 21
st
, 2023.
Copyright © 2023 Published by IRCS-ITB, ISSN: 2337-5787, DOI: 10.5614/itbj.ict.res.appl.2023.17.2.4
Generative Adversarial Networks Based Scene Generation
on Indian Driving Dataset
K. Aditya Shastry*, B.A. Manjunatha, M. Mohan, T.G. Mohan Kumar &
D.U. Karthik
Department of Information Science and Engineering,
Nitte Meenakshi Institute of Technology, Bengaluru, 560064, India
*E-mail: adityashastry.k@nmit.ac.in
Abstract. The rate of advancement in the field of artificial intelligence (AI) has
drastically increased over the past twenty years or so. From AI models that can
classify every object in an image to realistic chatbots, the signs of progress can be
found in all fields. This work focused on tackling a relatively new problem in the
current scenario-generative capabilities of AI. While the classification and
prediction models have matured and entered the mass market across the globe,
generation through AI is still in its initial stages. Generative tasks consist of an AI
model learning the features of a given input and using these learned values to
generate completely new output values that were not originally part of the input
dataset. The most common input type given to generative models are images. The
most popular architectures for generative models are autoencoders and generative
adversarial networks (GANs). Our study aimed to use GANs to generate realistic
images from a purely semantic representation of a scene. While our model can be
used on any kind of scene, we used the Indian Driving Dataset to train our model.
Through this work, we could arrive at answers to the following questions: (1) the
scope of GANs in interpreting and understanding textures and variables in
complex scenes; (2) the application of such a model in the field of gaming and
virtual reality; (3) the possible impact of generating realistic deep fakes on society.
Keywords: artificial intelligence; deep learning; driving dataset; generative adversial
networks; scene generation.
1 Introduction
The advent of artificial intelligence began in 1943 when the first paper on the
concept of a neural network was published. Progress in AI has come a long way,
from a theoretical neural network model to models that can detect, classify,
predict, and more recently, generate data they have been trained on. The internet
has been abuzz with deep fakes and the potential issues such technology could
cause with the advent of generative models. However, generative models provide
a twofold benefit: they can generate new data in areas lacking suitable data (such
as native language processing) and they can be used to interpret and analyze
already existing data.