Deep Learning for Billboard Classification

Sayali Chavan; Dermot Kerr; Sonya Coleman; Hussein Khader

Deep Learning for Billboard Classification

Sayali Chavan, Dermot Kerr, Sonya Coleman, Hussein Khader

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

87 Downloads (Pure)

Abstract

Advertising is essential to increase product awareness and foster a positive outlook, which in turn helps sales. To promote the brand and its products, billboard advertisements are widely used. This paper presents a novel approach for classifying billboards. The proposed method utilises Convolutional Neural
Network (CNN) architectures to extract features from the images to enable classification. The model is trained on a dataset of billboards collected from various locations and achieves results that demonstrate high classification accuracy. The system is trained and evaluated using the CIFAR10 dataset, which includes 10 classes of objects and an additional 11th class - ’billboard’,
is included. The experiment utilises five different CNN architectures: Basic CNN, ResNet, VGG, MobileNet, and DenseNet. The performance and evaluation of each architecture are presented in detail, and extensive experiments and comparisons are conducted to determine the most effective model for classifying billboards. The results indicate that a CNN and its architectural designs are a promising solution for automating the classification of billboards in the wild.

Original language	English
Title of host publication	DIGITAL 2023
Subtitle of host publication	Advances on Societal Digital Transformation
Editors	Petre Dini
Publisher	International Academy, Research, and Industry Association
Pages	29-35
Number of pages	7
ISBN (Print)	978-1-68558-115-2
Publication status	Published online - 11 Oct 2023
Event	Advances on Societal Digital Transformation DIGITAL 2023 - Portugal, Porto, Portugal Duration: 25 Sept 2023 → 29 Sept 2023 https://www.iaria.org/conferences2023/DIGITAL23.html

Conference

Conference	Advances on Societal Digital Transformation DIGITAL 2023
Abbreviated title	DIGITAL 2023
Country/Territory	Portugal
City	Porto
Period	25/09/23 → 29/09/23
Internet address	https://www.iaria.org/conferences2023/DIGITAL23.html

Keywords

Classification
CNN-Architecture
Billboard
Image-Processing
CIFAR10

Access to Document

Deep Learning for Billboard ClassificationAuthor Accepted Manuscript, 2.2 MB

Cite this

@inproceedings{8a0328019403498e931f6774b6c632eb,

title = "Deep Learning for Billboard Classification",

abstract = "Advertising is essential to increase product awareness and foster a positive outlook, which in turn helps sales. To promote the brand and its products, billboard advertisements are widely used. This paper presents a novel approach for classifying billboards. The proposed method utilises Convolutional NeuralNetwork (CNN) architectures to extract features from the images to enable classification. The model is trained on a dataset of billboards collected from various locations and achieves results that demonstrate high classification accuracy. The system is trained and evaluated using the CIFAR10 dataset, which includes 10 classes of objects and an additional 11th class - {\textquoteright}billboard{\textquoteright},is included. The experiment utilises five different CNN architectures: Basic CNN, ResNet, VGG, MobileNet, and DenseNet. The performance and evaluation of each architecture are presented in detail, and extensive experiments and comparisons are conducted to determine the most effective model for classifying billboards. The results indicate that a CNN and its architectural designs are a promising solution for automating the classification of billboards in the wild.",

keywords = "Classification, CNN-Architecture, Billboard, Image-Processing, CIFAR10",

author = "Sayali Chavan and Dermot Kerr and Sonya Coleman and Hussein Khader",

year = "2023",

month = oct,

day = "11",

language = "English",

isbn = "978-1-68558-115-2",

pages = "29--35",

editor = "Petre Dini",

booktitle = "DIGITAL 2023",

publisher = "International Academy, Research, and Industry Association",

address = "United States",

note = "Advances on Societal Digital Transformation DIGITAL 2023, DIGITAL 2023 ; Conference date: 25-09-2023 Through 29-09-2023",

url = "https://www.iaria.org/conferences2023/DIGITAL23.html",

}

TY - GEN

T1 - Deep Learning for Billboard Classification

AU - Chavan, Sayali

AU - Kerr, Dermot

AU - Coleman, Sonya

AU - Khader, Hussein

PY - 2023/10/11

Y1 - 2023/10/11

N2 - Advertising is essential to increase product awareness and foster a positive outlook, which in turn helps sales. To promote the brand and its products, billboard advertisements are widely used. This paper presents a novel approach for classifying billboards. The proposed method utilises Convolutional NeuralNetwork (CNN) architectures to extract features from the images to enable classification. The model is trained on a dataset of billboards collected from various locations and achieves results that demonstrate high classification accuracy. The system is trained and evaluated using the CIFAR10 dataset, which includes 10 classes of objects and an additional 11th class - ’billboard’,is included. The experiment utilises five different CNN architectures: Basic CNN, ResNet, VGG, MobileNet, and DenseNet. The performance and evaluation of each architecture are presented in detail, and extensive experiments and comparisons are conducted to determine the most effective model for classifying billboards. The results indicate that a CNN and its architectural designs are a promising solution for automating the classification of billboards in the wild.

AB - Advertising is essential to increase product awareness and foster a positive outlook, which in turn helps sales. To promote the brand and its products, billboard advertisements are widely used. This paper presents a novel approach for classifying billboards. The proposed method utilises Convolutional NeuralNetwork (CNN) architectures to extract features from the images to enable classification. The model is trained on a dataset of billboards collected from various locations and achieves results that demonstrate high classification accuracy. The system is trained and evaluated using the CIFAR10 dataset, which includes 10 classes of objects and an additional 11th class - ’billboard’,is included. The experiment utilises five different CNN architectures: Basic CNN, ResNet, VGG, MobileNet, and DenseNet. The performance and evaluation of each architecture are presented in detail, and extensive experiments and comparisons are conducted to determine the most effective model for classifying billboards. The results indicate that a CNN and its architectural designs are a promising solution for automating the classification of billboards in the wild.

KW - Classification

KW - CNN-Architecture

KW - Billboard

KW - Image-Processing

KW - CIFAR10

UR - https://www.iaria.org/conferences2023/CfPDIGITAL23.html

UR - https://www.thinkmind.org/index.php?view=article&articleid=digital_2023_1_50_20026

M3 - Conference contribution

SN - 978-1-68558-115-2

SP - 29

EP - 35

BT - DIGITAL 2023

A2 - Dini, Petre

PB - International Academy, Research, and Industry Association

T2 - Advances on Societal Digital Transformation DIGITAL 2023

Y2 - 25 September 2023 through 29 September 2023

ER -

Deep Learning for Billboard Classification

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint