TOY SORTING ALGORITHM USING CNN MODEL

A MINI PROJECT REPORT

18CSC305J - ARTIFICIAL INTELLIGENCE

Submitted by
VEERALLI HARSHAVARDHAN [RA2111056010023]

Under the guidance of

Dr. P.C. KARTHIK
Associate Professor, Department of Data Science and Business systems

in partial fulfillment for the award of the degree

of
BACHELOR OF TECHNOLOGY
in

COMPUTER SCIENCE & ENGINEERING [DATA SCIENCE]

of

FACULTY OF ENGINEERING AND TECHNOLOGY

S.R.M. Nagar, Kattankulathur, Chengalpattu District

MAY 2024
SRM INSTITUTE OF SCIENCE AND TECHNOLOGY
(Under Section 3 of UGC Act, 1956)

BONAFIDE CERTIFICATE

Certified that Mini project report titled “TOY SORTING ALGORITHM USING
CNN MODEL” is the bona fide work of VEERALLI HARSHAVARDHAN
[RA2111056010023] who carried out the minor project under my supervision.
Certified further, that to the best of my knowledge, the work reported herein does
not form any other project report or dissertation on the basis of which a degree or
award was conferred on an earlier occasion on this or any other candidate.

FACULTY SIGNATURE HOD SIGNATURE

DR. P.C. KARTHIK

Faculty of Engineering &
Technology, Kattankulathur -
Chennai
Department of Data Science And
Business Systems
 ABSTRACT

Our proposal entails the development of a rule-based sorting system explicitly designed for
children's toys. The primary objective of this system is to simplify toy organization tasks while
simultaneously providing children with a captivating and educational experience.

The sorting process itself revolves around categorizing toys into predetermined groups, guided
by specific rules determined by their attributes. These attributes encompass various aspects
such as size, shape, material composition, and functional features. By adhering to these rules,
the system ensures that each toy is assigned to the most appropriate category, facilitating
efficient organization.

To enhance the sorting process further, we aim to explore the integration of artificial
intelligence (AI) techniques, particularly machine learning algorithms. By leveraging AI, we
can automate and streamline the sorting process, leading to increased efficiency and scalability.
Machine learning algorithms will be trained on labeled datasets of toys, enabling them to
recognize patterns and attributes that dictate the categorization process.

The integration of AI techniques not only improves the efficiency of the sorting system but
also introduces an element of interactivity and engagement for children. Through interactive
interfaces and feedback mechanisms, children can actively participate in the sorting process,
learning about different attributes and their significance. This not only fosters organizational
skills but also stimulates curiosity and interest in technology from an early age.

Furthermore, by incorporating gamification elements and educational content, we can

transform the sorting activity into an enjoyable and enriching experience for children. By
making the sorting process fun and rewarding, we aim to motivate children to actively engage
with the system and develop valuable skills in organization, problem-solving, and critical
thinking.

ii
 TABLE OF CONTENTS

ABSTRACT ii

TABLE OF CONTENTS iii

ACKNOWLEDGEMENT iv

OWN WORK DECLARATION v

1 INTRODUCTION 1

2 LITERATURE SURVEY 2

3 SYSTEM ARCHITECTURE AND DESIGN 3

4 METHODOLOGY 4

5 CODING AND TESTING 5

6 SREENSHOTS AND RESULTS 6

7 CONCLUSION AND FUTURE ENHANCEMENT 8

8 REFERENCES 10

iii
 ACKNOWLEDGEMENTS

We express our humble gratitude to Dr C. Muthamizhchelvan, Vice-Chancellor, SRM Institute of Science

and Technology, for the facilities extended for the project work and his continued support.

We extend our sincere thanks to Dean-CET, SRM Institute of Science and Technology, Dr T.V.Gopal, for
his invaluable support.

We wish to thank Dr Revathi Venkataraman, Professor & Chairperson, School of Computing, SRM
Institute of Science and Technology, for her support throughout the project work.

We are incredibly grateful to our Head of the Department, Dr M. Lakshmi Professor, HoD, Department of
Data Science and Business Systems, SRM Institute of Science and Technology, for her suggestions and
encouragement at all the stages of the project work.

We register our immeasurable thanks to our Faculty Advisor, Dr.M.Arthy, Associate Professor, Department
of Data Science and Business Systems, SRM Institute of Science and Technology, for leading and helping us
to complete our course.

Our inexpressible respect and thanks to my guide also our Academic Advisor, Dr. P.C.Karthik , Associate
Professor, Department of Data Science and Business Systems, for providing me with an opportunity to
pursue my project under his mentorship. He provided us with the freedom and support to explore the research
topics of our interest. His passion for solving problems and making a difference in the world has always been
inspiring.

We sincerely thank the Data Science and Business Systems staff and students, SRM Institute of Science and
Technology, for their help during our project. Finally, we would like to thank parents, family members, and
friends for their unconditional love, constant support, and encouragement.

iv
 SRM Institute of Science and Technology

Own Work Declaration Form

Degree/Course : B.Tech CSE with specialization in Data Science

Student Details : SANJANA SENTHILKUMAR SATHYA [RA2111056010020]
P.ARAVIND [RA2111056010019]
TUHIN DUTTA [RA2111056010022]
VEERALLI HARSHA VARDHAN [RA2111056010023]

We hereby certify that this assessment compiles with the University’s Rules and Regulations
relating to Academic misconduct and plagiarism, as listed in the University Website, Regulations,
and the Education Committee guidelines. We confirm that all the work contained in this assessment
is our own except where indicated, and that we have met the following conditions:

• Clearly references / listed all sources as appropriate

• Referenced and put in inverted commas all quoted text (from books,

web, etc.)

• Not made any use of the report(s) or essay(s) of any other student(s)

either past or present

• Acknowledged in appropriate places any help that I have received from others (e.g. fellow

students, technicians, statisticians, external sources) • Compiled with any other plagiarism

criteria specified in the Course handbook / University website

• I understand that any false claim for this work will be penalized in accordance with the

University policies and regulations.

v
 CHAPTER 1

INTRODUCTION

Our current toy sorting system relies on predetermined criteria like size, shape, material, and
function to classify toys into different groups. The next step involves incorporating artificial
intelligence (AI) methods to automate this process, thereby improving efficiency and scalability.
This initiative is not solely about enhancing operational aspects; it's also about creating an
interactive and educational environment for children. By automating sorting tasks, we hope to
instill organizational skills and spark an early interest in technology among young learners.

Traditionally, sorting toys has been a manual and time-consuming task, requiring human
intervention to apply subjective rules. However, with the advent of AI and machine learning, we
have an opportunity to revolutionize this process. By leveraging labeled datasets and advanced
algorithms, we can train models to accurately categorize toys based on their attributes.

Our primary aim is to offer children an engaging and educational experience while simplifying the
sorting process. Through interactive tasks and real-time feedback, we intend to make sorting fun
and rewarding, encouraging children to actively participate. Additionally, by showcasing the role
of technology in everyday tasks, we hope to stimulate curiosity and foster a passion for learning.

The methodology involves collecting a diverse range of toys and attributes to develop a
comprehensive dataset. This dataset will serve as the foundation for training machine learning
models, which will then be integrated into the sorting system. A user-friendly interface will be
designed to facilitate interaction, ensuring that children can easily engage with the technology.

The anticipated outcomes of this project are manifold. Firstly, automating the sorting process will
significantly reduce time and effort, leading to enhanced efficiency. Moreover, by incorporating
educational elements, we aim to promote organizational skills and cultivate an early interest in
technology among children. Ultimately, this project represents a step towards creating a more
interactive and enriching learning environment for young learners.
 CHAPTER 2

LITERATURE SURVEY

1. Applications of CNNs

Long, J., Shelhamer, E., & Darrell, T. (2015). "Fully convolutional networks for semantic
segmentation." Proceedings of the IEEE conference on computer vision and pattern
recognition (CVPR), 3431-3440.

FCNs revolutionize semantic segmentation by enabling pixel-wise classification, leading

to significant advances in tasks like scene parsing and medical image segmentation.

2. Optimizations and Training Techniques

Ioffe, S., & Szegedy, C. (2015). "Batch normalization: Accelerating deep network training
by reducing internal covariate shift." Proceedings of the 32nd International Conference on
Machine Learning (ICML), 37, 448-456.

Batch normalization mitigates internal covariate shift during training, leading to faster
convergence and improved generalization performance.

3. Foundational Works

LeCun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). "Gradient-based learning applied
to document recognition." Proceedings of the IEEE, 86(11), 2278-2324.

This seminal paper introduces the LeNet-5 architecture, laying the groundwork for modern
CNNs and demonstrating their efficacy in handwritten digit recognition tasks.
 CHAPTER 3
SYSTEM ARCHITECTURE AND DESIGN
 CHAPTER 4

METHODOLOGY

To achieve our objectives, we plan to follow a systematic approach:

Data Collection: We will gather a diverse dataset of toys, encompassing a wide range of sizes,
shapes, materials, and functions. Each toy will be meticulously labeled with its corresponding
attributes.

Model Development: Using state-of-the-art machine learning techniques, we will develop a robust
classification model capable of accurately categorizing toys based on their attributes. This may
involve training convolutional neural networks (CNNs) or other deep learning architectures on the
collected dataset.

Integration with Sorting System: Once the model is trained and validated, we will integrate it
into the existing toy sorting system. This may involve deploying the model on embedded devices
or cloud platforms, depending on the scalability and performance requirements.

User Interface Design: To ensure a seamless user experience, we will design an intuitive user
interface for interacting with the sorting system. This interface will allow children to input toys
into the system and receive real-time feedback on the categorization process.
 CHAPTER 5

CODING AND TESTING

import tensorflow as tf
import numpy as np

# Create a toy dataset of images

images = np.random.rand(100, 28, 28, 1)

# Create labels for the images

labels = np.random.randint(0, 10, size=(100,))

# Create a CNN model

model = tf.keras.Sequential([
tf.keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)),
tf.keras.layers.MaxPooling2D((2, 2)),
tf.keras.layers.Conv2D(64, (3, 3), activation='relu'),
tf.keras.layers.MaxPooling2D((2, 2)),
tf.keras.layers.Flatten(),
tf.keras.layers.Dense(128, activation='relu'),
tf.keras.layers.Dense(10, activation='softmax')
])

# Compile the model

model.compile(
optimizer='adam',
loss='sparse_categorical_crossentropy',
metrics=['accuracy']
)

# Train the model

model.fit(images, labels, epochs=50)

# Sort the images by their predicted labels

predicted_labels = model.predict(images)
sorted_images = [images[i] for i in np.argsort(predicted_labels)]
 CHAPTER 6

SCREENSHOTS AND RESULTS

 CHAPTER 7

CONCLUSION AND FUTURE ENHANCEMENTS

In conclusion, the development of a rule-based sorting system for children's toys presents a
promising solution for facilitating toy organization while engaging children in an enjoyable and
educational activity. By leveraging predefined rules based on attributes such as size, shape,
material, and function, the system streamlines the sorting process, ensuring that each toy is
categorized accurately and efficiently.

Furthermore, the integration of artificial intelligence (AI) techniques, particularly machine learning
algorithms, enhances the system's efficiency and scalability. By automating the sorting process,
the system not only reduces the time and effort required for organization but also introduces an
interactive and engaging element for children. Through interactive interfaces and feedback
mechanisms, children can actively participate in the sorting process, fostering organizational skills
and promoting an early interest in technology.

Overall, the rule-based sorting system represents a significant step towards creating a more
efficient and enjoyable toy organization experience for both children and caregivers. By combining
innovative technology with educational objectives, the system not only simplifies everyday tasks
but also inspires and educates young minds, laying the foundation for lifelong learning and
development.

Future Enhancements:

Looking ahead, several enhancements and advancements can be considered to further improve the
rule-based sorting system:

Refinement of Rules: Continuously refine and optimize the predefined rules based on user
feedback and data analysis to ensure accurate and efficient toy categorization.
Integration of Advanced AI Techniques: Explore advanced AI techniques, such as deep learning
algorithms, for more sophisticated toy recognition and categorization, potentially improving the
system's accuracy and adaptability.
Personalization: Implement personalized sorting profiles for individual users, allowing children
to customize their sorting preferences and interact with the system in a more tailored manner.

Enhanced Interactivity: Introduce additional interactive features and gamification elements to

make the sorting process even more engaging and enjoyable for children, further encouraging
active participation.

Expansion of Educational Content: Incorporate additional educational content and learning

modules into the system to provide children with opportunities for skill development and
knowledge acquisition beyond toy organization.

Integration with Smart Devices: Explore integration with smart devices and IoT technologies to
enable seamless communication and interaction between the sorting system and other connected
devices or applications.

Accessibility Features: Implement accessibility features to ensure inclusivity and accommodate

children with diverse needs and abilities, making the sorting system accessible to a broader
audience.

By continuously innovating and iterating upon the rule-based sorting system, we can further
enhance its effectiveness, usability, and educational value, ultimately providing children with a
more enriching and rewarding experience while promoting organization skills and fostering an
early interest in technology.
 CHAPTER 8

REFERENCES

1. https://www.analyticsvidhya.com/blog/2021/05/convolutional-neural-networks-cnn/

2. https://www.geeksforgeeks.org/convolutional-neural-network-cnn-in-machine-learning/

3. https://medium.com/@m.fosci/a-shape-detection-toy-experiment-for-convolutional-neural-
networks-50af2b191d21

4. https://www.kernix.com/article/a-toy-convolutional-neural-network-for-image-classification-
with-keras/