BOOK RECOMMENDATION SYSTEM

A PROJECT REPORT

Submitted by

M. KRISHNA VAMSI RA2211003020179

SANA LOKESH REDDY RA2211003020172
BANDI VENKATA SIDDHARTHA
RA2211003020144
P. AJAY KUMAR RA2211003020143

Under the guidance of

Dr. T. TAMIL SELVI

(Professor, Department of Computer Science and
Engineering)

in partial fulfilment for the award of the

degree of

BACHELOR OF TECHNOLOGY
in

COMPUTER SCIENCE AND ENGINEERING

of

COLLEGE OF ENGINEERING AND TECHNOLOGY

SRM INSTITUTE OF SCIENCE AND TECHNOLOGY

RAMAPURAM, CHENNAI -600089
OCT 2024
 SRM INSTITUTE OF SCIENCE AND TECHNOLOGY
(Deemed to be University U/S 3 of UGC Act, 1956)

BONAFIDE CERTIFICATE

Certified that this project report titled “BOOK RECOMMENDATION SYSTEM” is the
bonafide work of M. KRISHNA VAMSI (RA2211003020179), SANA LOKESH REDDY
(RA2211003020172), BANDI VENKATA SIDDHARTHA(RA2211003020144), P. AJAY
KUMAR (RA2211003020143) who carried out the project work 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 occasion on this or any other candidate.

SIGNATURE SIGNATURE

Dr. T. TAMIL SELVI Dr. K. RAJA, M.E., Ph.D.,

Professor Professor and Head
Computer Science and Engineering, Computer Science and Engineering,
SRM Institute of Science and Technology, SRM Institute of Science and Technology,
Ramapuram, Chennai. Ramapuram, Chennai.

Submitted for the project viva-voce held on ___________ at SRM Institute of Science
and Technology, Ramapuram, Chennai -600089.

INTERNAL EXAMINER 1 INTERNAL EXAMINER 2

 SRM INSTITUTE OF SCIENCE AND TECHNOLOGY
RAMAPURAM, CHENNAI - 89

DECLARATION

We hereby declare that the entire work contained in this project report titled “BOOK
RECOMMENDATION SYSTEM” has been carried out by M. KRISHNA VAMSI
(RA2211003020179), SANA LOKESH REDDY (RA2211003020172), BANDI
VENKATA SIDDHARTHA (RA2211003020144), P. AJAY KUMAR
(RA2211003020143) at SRM Institute of Science and Technology, Ramapuram Campus,
Chennai- 600089, under the guidance of Dr. T. TAMIL SELVI Professor, Department of
Computer Science and Engineering.

Place: Chennai
Date:

M. KRISHNA VAMSI

SANA LOKESH REDDY

BANDI VENKATA SIDDHARTHA

P. AJAY KUMAR

4
 ACKNOWLEDGEMENT

We place on record our deep sense of gratitude to our lionized Chairman Dr. R.

SHIVAKUMAR for providing us with the requisite infrastructure throughout the course.

We take the opportunity to extend our hearty and sincere thanks to our Dean, Dr. M.

MURALI KRISHNA, B.E., M. Tech., Ph.D. MISTE, FIE, C. Engg, for maneuvering us

into accomplishing the project.

We take the privilege to extend our hearty and sincere gratitude to the Professor and Head

of the Department, Dr. K. RAJA, M.E., Ph.D., for his suggestions, support and

encouragement towards the completion of the project with perfection.

We express our hearty and sincere thanks to our guide Dr. T. TAMIL SELVI, Professor,

Computer Science and Engineering Department for her encouragement, consecutive

criticism and constant guidance throughout this project work.

Our thanks to the teaching and non-teaching staff of the Computer Science and

Engineering Department of SRM Institute of Science and Technology, Ramapuram

Campus, for providing necessary resources for our project.

5
 Department of Computer Science and Engineering
SRM Institute of Science and Technology
Own Work Declaration form
To be completed by the student for all assessments

Degree/ Course : B. Tech Computer Science and Engineering

Student Name : M. KRISHNA VAMSI, SANA LOKESH REDDY, BANDI
VENKATA SIDDHARTHA, P. AJAY KUMAR
Registration Number: RA2211003020179, RA2211003020172, RA2211003020144,
RA2211003020143
Title of Work: BOOK RECOMMENDATION SYSTEM
I / We hereby certify that this assessment complies 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.
I / We confirm that all the work contained in this assessment is my / our own except
where indicated, and that I / 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)
● Given the sources of all pictures, data etc. that are not my own
● 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.

DECLARATION:
I am aware of and understand the University’s policy on Academic misconduct and
plagiarism and I certify that this assessment is my / our own work, except were
indicated by referring, and that I have followed the good academic practices noted
above.
If you are working in a group, please write your registration numbers and sign with the
date for every student in your group.

RA2211003020179 RA2211003020172 RA2211003020144

RA2211003020143

6
 ABSTRACT

The World Wide Web has become an immense repository of information, with commercial
activity increasing daily as hundreds of new companies add web pages. While this
abundance of content provides users with vast choices, it has also led to the issue of
information overload, where individuals struggle to find relevant information amidst the
flood of available data. Recommender systems have emerged as a solution to this problem
by offering personalized suggestions that help users discover content aligned with their
interests. These systems typically rely on community opinions, user behavior, or
preferences to make recommendations. This thesis aimed to design and evaluate several
approaches for generating personalized book recommendations. The process began with a
thorough investigation of existing recommender systems and user profiling techniques.
Understanding these foundational elements was essential to develop effective methods for
capturing user preferences and behavior. User profiles were constructed by monitoring
behavior patterns such as book selection, ratings, and browsing history. Based on this data,
three different approaches to generating book recommendations were developed: content-
based filtering, collaborative filtering, and a hybrid model combining the two. The final
phase of the project involved an evaluation of recommendation accuracy. This was done
using both live user experiments and offline analysis, employing appropriate testing
methodologies to assess the precision of the recommendations. Live experiments allowed
for direct feedback from users regarding the relevance of the suggestions, while offline
analysis provided a more structured evaluation of the system's performance across various
metrics such as precision and recall. The results of the evaluation were promising,
indicating that the system consistently produced highly accurate recommendations. The
hybrid approach, which integrates both content-based and collaborative filtering, proved to
be the most effective, offering a superior balance between personal preferences and the
insights gathered from a community of users. This combination resulted in more accurate
and diverse book suggestions, effectively addressing the challenge of information overload
within the book domain and enhancing the overall user experience.

7
 TABLE OF CONTENTS
ABSTRACT iv

S.N TITLE Page No.

O
1. INTRODUCTION 1
2. LITERATURE SURVEY 6
3. PROPOSED METHODOLOGY 9
4. RESULTS AND DISCUSSION 13
5. CONCLUSION 18
6. FUTURE ENHANCEMENTS 21
7. SOURCE CODE 26
REFERENCES 37

8
 Chapter 1

INTRODUCTION

1.1 Introduction
This is an experimental project which first, designs…. And second evaluates different
approaches for offering recommendations to readers regarding books they may wish to
purchase, as part of an online bookshop website. Today the World Wide Web has provided
access to a vast array of information through the web pages, as a result of the Internet
growth. Also, commercial activity on the Web has increased to the point where hundreds
of new companies are adding Web pages daily. With this increase in the information
sources, a problem of information overload occurs, in which the users are trying to deal
with an excess of information that is not useful to them as they try to make sensible
decisions (Losee, 1989). As a response to this problem, a range of tools to help with
retrieving, searching, and filtering have been developed. The tool most widely used to
alleviate the problem of information overload is the search engine. The benefits for the
users from search engine technology have decreased as the number of web pages has
grown. In addition, the user must first consider the large number of search tools available
and decide which one to access. Then the user must interact with each one individually
because search engines are typically not personalized to individual users or their prevailing
context. Users usually make a choice on the basis of their personal experience or other
people’s experience. Based on these facts, recommender systems have been developed to
provide recommendations that help individual users identify content of interest by using
the opinions of a community of users and/or the user’s preferences. Today various
recommendation systems play an important role in supporting commercial websites to help
users find items that they know they would like to purchase, as well as discover new items
about which they had been unaware. The ability to persuade the consumers to buy a
suitable item is a significant goal for any recommender system in an ecommerce
environment. However, for any recommender system to be successful, the consumer must
trust and accept the 2 system’s recommendations. This is done with a clear explanation
from the system, presented in a way that is in keeping with the consumer’s preferences. A
good recommender system can significantly contribute to achieving the consumer’s
acceptance of the system recommendations.
9
 Chapter 2

LITERATURE REVIEW

An idea for Book Recommendation System Using ML Tools for Text Categorization was
put forth by Raymond J. Mooney and Loriene Roy. They outline a content-based book
recommendation system that classifies text using machine learning and information
extraction. Item recommendations are made based on information about the item itself
rather than on the preferences of other users. On the other hand, learning customized
profiles from descriptions of examples enables a system to uniquely characterize each
customer without the need to match his or her interests to another's. Through the use of
automatic text-categorization techniques on semi-structured text downloaded from the
web, they have been investigating content-based book recommendations. A database of
book data is used by the present prototype system, called LIBRA (Learning Intelligent
Book Recommending Agent). After employing a Bayesian learning algorithm to learn
about the user's profile, the system generates a ranked list of the most highly suggested
additional books from its library. Even when the method provides very modest training
sets, the overall results are quite positive. An original content-based book recommender
called LIBRA makes use of a straightforward. A unique book recommendation system was
proposed by Binge Cui and Xin Chen. When readers are unable to locate the desired book
using the library's bibliographic retrieval system, they are directed to the recommendation
pages. It is a web-based system for recommending books to a library's patrons. After
logging in, a user can search for books using author names or keywords like book titles. A
bibliographic retrieval system will then look for books using the same keywords. If the
recommendation system returns any results, submit these keywords to the web books
retrieval module. Web Books Retrieval Module allows the librarian or administrator of the
online book recommendation system to search the online bookshop using keywords by
creating accounts on sites like Amazon. As a result, the web retrieval module searches
these online bookshops as the logged-in user when the keyword is presented to it. The user
will receive the results from these online booksellers in the form of recommendations. The
statistic and analysis 4 module will determine the value of that specific book based on user
10
recommendations. The Auto-Order Module will then generate a book order automatically
based on the analysis results according to this value of book. The Short Message and Email
Notification Module will get a report from the Book Storage System once the purchased
books have been shelved. Then, utilizing a message and email server, it will inform the
readers who have suggested the books that have been acquired
Basic Statistics An effective and best-in-class hybrid recommendation algorithm was
proposed by Dharmendra Pathak, Sandeep Matharia, and C. N. S. Murthy, and it provides
recommendations that are more in line with user preferences. The hybrid recommendation
in this case combines collaborative, content-based, and context-based algorithms. The
primary input for collaborative filtering is rating, or the votes of numerous users, content-
based data, which is user-specific information like interests, dates of birth, and priorities,
and context-based data, which is behavioral information like time, taste, mood, and
weather. The similarity is measured using the cosine similarity. According to the user's
prior history, there are subject priority. When they buy a book, do they check to see if the
subject priority is changed from what was previously set? If so, topic priority 3 and subject
priority 2 should be reset. Priority 1 will remain the same. They came to the conclusion
that the proposed Hybrid book recommendation algorithm is superior to the others based
on 5 calculations and results. A book recommendation system based on integrated features
of content filtering, collaborative filtering, and association rule mining was proposed by
Anand Shanker Tewari, Abhay Kumar, and Asim Gopal Barman. When a customer
searches for a book, the search is recorded in the customer's purchase or search history.
Recommendation services conduct some filtering when a customer is not online, and the
results are saved in the customer's web profile. The recommendations will be created
automatically the following time the customer visits the website. Web Usage Mining
(WUM) is used in content-based filtering to give customers the pertinent information they
need. web server access logs, browser caches, or proxy logs are some examples of
historical data that web Usage Mining (WUM) commonly uses to extract knowledge. web
use Internet user behavior is recorded by mining, which then processes the information.
The item-based collaborative recommendation algorithm and cosine similarity are both
used to measure similarity. Compare the outcomes of collaborative content filtering with
association rule mining.

11
 Chapter 3
PROPOSED METHODOLOGY

The proposed methodology for the book recommendation system is designed to deliver
highly personalized, intuitive, and accurate recommendations by integrating advanced
algorithms with a deep understanding of user preferences. The process begins with the
creation of detailed user profiles. This is achieved by collecting data based on users'
reading behaviors—tracking their reading history, ratings, reviews, and browsing patterns.
Each interaction provides critical insights into the user's unique preferences, and this
process is dynamic. As users continue to engage with the system, their profiles are
constantly updated, reflecting their evolving tastes over time. This allows the system to
offer recommendations that are always aligned with the current interests of the reader. At
the same time, the system gathers comprehensive information about the books in its
database. This includes traditional metadata such as titles, authors, genres, and
descriptions, but it also goes beyond these basics. By leveraging Natural Language
Processing (NLP), the system analyzes the semantic content of the books to understand
deeper elements like themes, tone, and writing style. This additional layer of analysis
ensures that the system can make more nuanced recommendations, offering users
suggestions that reflect not just what the books are about, but how they feel and read,
catering to more personalized tastes. The recommendation engine is built upon two
fundamental approaches: collaborative filtering and content-based filtering. Collaborative
filtering looks at patterns in user behavior across the system, identifying commonalities
among readers with similar tastes. For example, if two users consistently enjoy similar
books, the system will recommend a book enjoyed by one to the other. Content-based
filtering, on the other hand, focuses on the specific attributes of books that a user has
enjoyed in the past, such as genre, themes, or writing style, and recommends similar books
accordingly. This ensures that the system can generate recommendations that are highly
personalized to the individual user.By combining both collaborative and content-based
filtering approaches into a hybrid model, the system achieves a balance between
personalized recommendations and broader trends across the user base. This hybrid model
allows for a more diverse set of suggestions, offering books that align with the user’s
personal preferences while also introducing new options based on the reading habits of
similar users.The system's performance is evaluated through live user experiments and
offline analysis, using metrics like precision, recall, and user satisfaction to measure
recommendation accuracy. Feedback is continuously incorporated, enabling the system to

12
adapt and improve over time. This human-centered approach ensures the recommendation
system provides not only technically accurate results but also an engaging and personalized
reading experience that evolves with the user input.

Advantages

 Personalized Experience: The system tailors’ recommendations to each user's unique

preferences, ensuring they discover books that align with their interests and reading habits.
This personalized approach makes it easier for users to find books they’re likely to enjoy,
enhancing their overall reading experience.

 Reduces Information Overload: With an overwhelming number of books available

online, users often struggle to choose the right one. A recommendation system filters
through this vast selection, presenting only the most relevant options and helping users
avoid choice paralysis.

 Time Efficiency: The system saves users time by quickly offering curated suggestions
based on their previous interactions, allowing them to focus more on reading and less on
searching for their next book.

 Improves User Engagement: By continuously learning from user feedback and

behavior, the system keeps recommendations fresh and relevant. This ongoing adaptation
leads to higher user satisfaction and encourages continued engagement with the platform.

 Diverse Discovery: The combination of collaborative and content-based filtering

introduces users to books they might not have considered but are likely to enjoy. This
expands their reading horizons and helps them explore new genres or authors.

 Adaptability and Scalability: The system’s ability to scale with large datasets and
adapt to changing user preferences makes it highly effective in real-time, ensuring that
even as user tastes evolve, the system remains accurate and reliable.

General Architecture:

13
 Figure 3.1: Architecture Diagram

Figure 3.1 The architecture comprises a data collection layer that gathers user interactions
and book metadata. A processing layer uses algorithms like collaborative filtering and
content-based filtering to analyze this data. The recommendation engine generates
personalized suggestions, while the feedback loop allows continuous learning. A user
interface displays the recommendations.

Figure 3.2 User data (ratings, history) and book metadata are collected and flow into the
data storage. This data is processed through filtering algorithms (collaborative, content-
based), which generate recommendations. These recommendations are sent to the user
interface for display, with user feedback looping back into the system to refine future
suggestions..

14
Figure 3.2: Data Flow Diagram

15
 Chapter 4

RESULTS AND DISCUSSION

4.1Input and Output

Figure 4.1: Output 1

Figure 4.2: Output

Figure 4.3: Output 3

16
4.2Efficiency of Proposed System
The efficiency of the proposed system stems from its ability to streamline feature
extraction and classification processes while maintaining high accuracy in image
segmentation tasks. Here's a breakdown of the context:
 Efficient Feature Fusion: The proposed system incorporates a cross-fusion module that
expands the perceptual field of low-level feature maps. By encoding more accurate
semantic information from small-scale objects, this module ensures that relevant
features are efficiently captured, minimizing redundancy and maximizing the
utilization of computational resources.
 Class-Specific Feature Extraction: Unlike traditional methods that may struggle with
complex and overlapping classes, the proposed system effectively captures
discriminative class-specific features. This targeted approach reduces computational
overhead by focusing resources on relevant features, improving both efficiency and
accuracy in classification tasks.
 Automatic Channel Selection: The system's cross-layer connectivity allows for the
automatic selection of important channels for feature learning. This adaptive
mechanism optimizes resource allocation during training, resulting in faster
convergence and reduced training times without compromising on model performance.
 Streamlined Feature Representation: By extracting feature information during encoding
and decoding splicing, the proposed system minimizes the loss of complex semantic
information. This streamlined approach ensures that relevant features are preserved
throughout the segmentation process, enhancing efficiency and accuracy in identifying
regions of interest.
 Real-world Applicability: The proposed system's efficiency extends beyond
computational performance to practical utility. Its ability to suppress characteristic
responses of irrelevant background areas and capture multiscale information makes it
well-suited for a wide range of real-world applications, from medical image analysis to
industrial inspection tasks.
Overall, the proposed system's efficiency lies in its ability to optimize feature
extraction, classification, and representation processes, leading to faster inference times,
reduced computational complexity, and improved performance across various
segmentation tasks.
4.3 Comparison between Existing and proposed System

17
 The existing system, while striving to enhance image segmentation, particularly
struggles with aligning feature maps of different resolutions, leading to increased model
complexity and inefficiency in extracting crucial feature insights. Moreover, it falls short in
effectively capturing both local and global features, particularly in images with fuzzy
features, which significantly reduces accuracy in segment recognition. Additionally, the
existing system may fail to ensure continuity and closure in extracted contours, limiting
their effectiveness, and requires careful tuning of several parameters in the network.
In contrast, the proposed system presents several advantages. It expands the perceptual
field of low-level feature maps through a cross-fusion module, enhancing semantic
information extraction from small-scale objects. By capturing discriminative class-specific
features and handling complex and overlapping classes, it demonstrates superior
performance and generalization capabilities.
The proposed system's cross-layer connectivity facilitates automatic selection of
important channels for feature learning, resulting in more effective feature representation
and faster training speed. Furthermore, it extracts feature information while encoding and
decoding splicing, mitigating loss of complex semantic information. Overall, the proposed
system not only aligns with modern developments but also significantly improves image
segmentation accuracy, presenting a fresh and enhanced approach to the subject. It
effectively suppresses characteristic responses of irrelevant background areas, captures
multiscale information, and offers a promising solution for accurate segmentation.
In summary, while the existing system faces challenges related to feature alignment,
feature extraction, and contour continuity, the proposed system presents a comprehensive
approach with advanced feature fusion, class-specific feature extraction, and enhanced
performance metrics. The proposed system's advantages signify a substantial leap forward
in image segmentation accuracy and effectiveness compared to the limitations of the
existing system.
4.4Results

18
Figure 4.4: Training Accuracy and Validation Accuracy

Figu
re 4.5: Training Loss and Validation Loss

19
Figure 4.6: Training Jaccard and Validation Jaccard Coefficient

20
 Chapter 5

CONCLUSION
In this paper, we introduce a groundbreaking neural network-based approach for
semantic segmentation of indoor scenes that incorporates both HHA and RGB images to
increase segmentation reliability. Our novel Visual Optimization Learning Model
significantly enhances the learning process by improving feature representation,
robustness, and discrimination, while also effectively utilizing cross-modal features. The
model features an encoder-decoder architecture with a dual-branch encoder dedicated to
extracting and combining unique RGB image features. Focused on addressing the
challenges of semantic segmentation in complex backgrounds, our study specifically
tackles the discord between semantic response values and object semantics critical to real-
time machine vision detection.
We conducted extensive experiments that demonstrate the effectiveness of our model,
comparing it with leading semantic segmentation models across various metrics, including
global accuracy, mean accuracy, and mean intersection over union. Based on our findings,
we recommend the Visual Optimization Learning Model as the most suitable for fire
detection and segmentation tasks, highlighting its superiority in handling complex
semantic scenarios.

21
 Chapter 6

FUTURE ENHANCEMENT

In our upcoming research, we intend to improve our network's efficiency by reducing

its time and space requirements. Furthermore, we will investigate semi-supervised or
weakly-supervised learning methods for indoor-scene semantic segmentation to address
the limitations of small datasets and the challenges of inaccurate data labeling.

22
 Chapter 7

SOURCE CODE
from flask import Flask,render_template,request
import pickle
import numpy as np

popular_df = pickle.load(open('popular.pkl','rb'))
pt = pickle.load(open('pt.pkl','rb'))
books = pickle.load(open('books.pkl','rb'))
similarity_scores = pickle.load(open('similarity_scores.pkl','rb'))

app = Flask(_name_)

@app.route('/')
def index():
return render_template('index.html',
book_name = list(popular_df['Book-Title'].values),
author=list(popular_df['Book-Author'].values),
image=list(popular_df['Image-URL-M'].values),
votes=list(popular_df['num_ratings'].values),
rating=list(popular_df['avg_rating'].values)
)

@app.route('/recommend')
def recommend_ui():
return render_template('recommend.html')

@app.route('/recommend_books',methods=['post'])
@app.route('/recommend_books', methods=['POST'])
@app.route('/recommend_books', methods=['POST'])
def recommend():
user_input = request.form.get('user_input')

result = np.where(pt.index == user_input)[0]

if result.size == 0:
return render_template('recommend.html', data=[], message="Book not found. Please
try a different title.")

index = result[0]
similar_items = sorted(list(enumerate(similarity_scores[index])), key=lambda x: x[1],
reverse=True)[1:5]

data = []
for i in similar_items:
item = []

23
 temp_df = books[books['Book-Title'] == pt.index[i[0]]]
item.extend(list(temp_df.drop_duplicates('Book-Title')['Book-Title'].values))
item.extend(list(temp_df.drop_duplicates('Book-Title')['Book-Author'].values))
item.extend(list(temp_df.drop_duplicates('Book-Title')['Image-URL-M'].values))

data.append(item)

return render_template('recommend.html', data=data, message=None)

if _name_ == '_main_':
app.run(debug=True)

24
 References

[1] Chen Yang, Qi Chen, Yaoyao Yang, Jingyu Zhang, Minshun Wu, Kuizhi Mei, “SDF-
SLAM: A Deep Learning Based Highly Accurate SLAM Using Monocular Camera
Aiming at Indoor Map Reconstruction With Semantic and Depth Fusion”, IEEE Access,
2022.
[2] Lijie Yu, Yuliang Sun, Xudong Zhang, Yongwei Miao, Xiuli Zhang, “Point Cloud
Instance Segmentation of Indoor Scenes Using Learned Pairwise Patch Relations”, IEEE
Access, 2021.
[3] Usman Ahmad Usmani, Junzo Watada, Jafreezal Jaafar, Izzatdin Abdul Aziz, Arunava
Roy, “A Reinforced Active Learning Algorithm for Semantic Segmentation in Complex
Imaging”, IEEE Access, 2021.
[4] Elhassan Mohamed, Konstantinos Sirlantzis, Gareth Howells, “Indoor/Outdoor Semantic
Segmentation Using Deep Learning for Visually Impaired Wheelchair Users”, IEEE
Access, 2021.
[5] Xinyang Zhao, Changhong Wang, Marcelo H. Ang, “Real-Time Visual-Inertial
Localization Using Semantic Segmentation Towards Dynamic Environments”, IEEE
Access, 2020
[6] Yanran Wang, Qingliang Chen, Shilang Chen, Junjun Wu, “Multi-Scale Convolutional
Features Network for Semantic Segmentation in Indoor Scenes”, IEEE Access, 2020
[7] Chongben Tao, Zhen Gao, Jinli Yan, Chunguang Li, Guozeng Cui, “Indoor 3D Semantic
Robot VSLAM Based on Mask Regional Convolutional Neural Network”, IEEE Access,
2020.
[8] Shuangquan Han, Zhihong Xi, “Dynamic Scene Semantics SLAM Based on Semantic
Segmentation”, IEEE Access, 2020.
[9] Chen Wang, Yue Qi, “A Real-Time Indoor Scene Analysis Method Based on RGBD
Stream”, IEEE Access, 2019.
[10] V. Badrinarayanan, A. Kendall and R. Cipolla, "SegNet: A Deep Convolutional Encoder-
Decoder Architecture for Image Segmentation," in IEEE Transactions on Pattern Analysis
and Machine Intelligence, 2017.

25
SRM INSTITUTE OF SCIENCE AND TECHNOLOGY
(Deemed to be University u / s 3 of UGC Act, 1956)

Office of Controller of Examinations

REPORT FOR PLAGIARISM CHECK ON THE DISSERTATION / PROJECT REPORT FOR UG /
PG PROGRAMMES
(To be attached in the dissertation / project report)
KANISHK V
1 Name of the Candidate (IN BLOCK
LETTERS) VINEESH K
PRAVEEN RAJ A

Bharathi Salai, Ramapuram, Chennai-89.

2 Address of Candidate Mobile Number: 9884983488, 9962241441,
9003249502.

3 Registration Number RA2011026020093, RA2011026020120,

RA2011026020124

4 Date of Birth 26/10/2002, 15/12/2002, 12/01/2003

Computer Science and Engineering with

5 Department Specialization in Artificial Intelligence and
Machine Learning

6 Faculty Faculty of Engineering and Technology

Pattern-Aware Indoor Semantic Segmentation

7 Title of the Dissertation / Project based on Visual Optimization Learning

Individual or Group: Group (Strike whichever is

not applicable)
a) If the project / dissertation is done in group,
then how many students together completed
8 Whether the above project / dissertation the project: 03
is done by b) Mention the Name and Register number of
other candidates:
KANISHK V [RA2011026020093]
VINEESH K [RA2011026020120]
PRAVEEN RAJ A [RA2011026020124]

Dr. A. SENTHILSELVI, Professor, Department

of Computer Science and Engineering, SRM
Institute of Science and Technology,
9 Name and address of the Supervisor / Ramapuram Campus, Chennai 89
Guide
Mail ID : [email protected]
Mobile Number : 95789 20861

26
 NA
10 Name and address of the Co-Supervisor
/ Guide Mail ID: NA Mobile Number: NA

Turnitin
11 Software Used

12 Date of Verification 30/04/2024

13 Plagiarism Details: (to attach the final report from the software)

Percentage of Percentage of % of plagiarism

similarity index similarity after excluding
Chapter Title of the Report (including self index Quotes,
citation) (Excluding Bibliography,
self citation) etc.,

Towards Intelligent Deep Enhanced

1 Feature Representation for Image NA NA 6%
Denoising using Encoder-Decoder
Algorithm
Appendices NA NA NA

I / We declare that the above information has been verified and found true to the best of my / our
knowledge.

Name and Signature of the Staff

Signature of the Candidate (Who uses the plagiarism check software)

Name and Signature of the Supervisor / Guide Name and Signature of the Co-Supervisor / Co-
Guide

Dr. K. Raja
Name and Signature of the HOD

27
28
29