MINISTRY OF EDUCATION AND TRAINING HO CHI MINH CITY UNIVERSITY OF TECHNOLOGY AND EDUCATION GRADUATION THESIS INFORMATION TECHNOLOGY CLASSIFYING THE DESIGN STYLE OF AN INTERIOR OBJECT USING DEEP LEARNING INSTRUCTOR: Ms. MAI ANH THO STUDENT: BIEN QUANG HUY SKL013463 Ho Chi Minh City, 2024 MINISTRY OF EDUCATION AND TRAINING HO CHI MINH UNIVERSITY OF TECHNOLOGY AND EDUCATION FACULTY FOR HIGH-QUALITY TRAINING PROJECT IN SOFTWARE ENGINEERING CLASSIFYING THE DESIGN STYLE OF AN INTERIOR OBJECT USING DEEP LEARNING Advisor: Ms. Mai Anh Thơ Students: Biện Quang Huy 19110005 HO CHI MINH CITY - 2024 CỘNG HOÀ XÃ HỘI CHỦ NGHĨA VIỆT NAM Độc lập – Tự do – Hạnh phúc ……***…… Ho Chi Minh, Day…Month…Year 2024 SPECIALIZED ESSAY MISSION Full name: Biện Quang Huy Student ID: 19110005 Class: 19110CLA2 Major: Software Engineering Instructor guide: Ms. Mai Anh Thơ Date of project receipt: 1/3/2024 Project submission date: 14/6/2024 1.
Project Name: Classifying the design style of an interior object using deep learning. Data, research paper: Datasets from instructor guide, research papers in Deep Learning, CNN models from instructor and self-search. Project contents: - Building prediction model based on Interior static images using ResNet50, VGG19, InceptionV3, Xception. - Building an intuitive website for Interior image classification 4.
Products: A website for design style of an interior object classification 2 CHAIR OF THE PROGRAM INSTRUCTOR 3 CỘNG HOÀ XÃ HỘI CHỦ NGHĨA VIỆT NAM Độc lập – Tự do – Hạnh phúc ……***…… Ho Chi Minh, Day…Month…Year 2024 INSTRUCTOR COMMENTARY Full name: Biện Quang Huy Student ID: 19110005 Class: 19110CLA2 Major: Software Engineering Project name: Classifying the design style of an interior object using deep learning Reviewer teacher: Ms. Mai Anh Thơ COMMENTARY 1. Topic content and workload Theory o Study and research about design style of an interior object o Research about machine learning and deep learning: ResNet50, VGG19, InceptionV3, Xception. o Propose a method of applying deep learning to solve the problem of design style of interior object classification Experiments o Using YOLOV5 for object segmentation to collect a dataset of interior objects labeling on design styles.
o Implement 04 models for classifying types of design style of interior object: ResNet50, VGG19, InceptionV3, Xception with the accuracy of 90%. o Build an application using the above model on a simple website. Links 4 o https://github.com/Hb011/capstonee 2. Advantages The interior object dataset can be reused.
The models have been trained with high accuracy. The website is friendly and easy to use. Students have the capability of self-studying and active learning. The project can be accomplished as scheduled.
Disadvantages Limited time for developing project. The models have not been modified. Recommend for protection or not?. ) Ho Chi Minh, June 14th ,2024 Instructor’s commentary 5 CỘNG HOÀ XÃ HỘI CHỦ NGHĨA VIỆT NAM Độc lập – Tự do – Hạnh phúc ……***…… Ho Chi Minh, Day…Month…Year 2024 REVIEWER'S COMMENTARY Full name: Biện Quang Huy Student ID: 19110005 Class: 19110CLA2 Major: Software Engineering Project name: Classifying the design style of an interior object using deep learning Reviewer teacher: Mr.
Quách Đình Hoàng COMMENTARY 1. Regarding the content of the topic and the volume of implementation:. Recommend for defending or not? …. 6 Ho Chi Minh, June 14th ,2023 Reviewer’s commentary 7 ACKNOWLEDGEMENT I would want to express my sincere gratitude to the Department of High-Quality Training, Ho Chi Minh City University of Technology and Education, and all the professors and teachers who have actively taught and aided me during my study and research.
In particular, I would need to specific my true appreciation to Ms. Mai Anh Tho for personally instructing, leading and establishing all suitable conditions to let me carry out the thesis. She helped us a lot in picking a topic and directing us to understand, investigate the theory and practice the issue. During the thesis implementation, the knowledge that the teacher provided us was extremely valuable, not only assisting us in completing the thesis but also supplementing us with a significant amount of core knowledge.
However, due to our limited professional expertise and limitations in our personal experience, the substance of the report is unavoidable. I would like to get your thoughts and more direction. She guides and criticizes to make this report more thorough. Ho Chi Minh city, June 14th, 2024 Biện Quang Huy 8 PREFACE Computer technologies are becoming increasingly popular as the information technology era progresses, and they play an important role in human life.
Machines are gradually replacing humans in heavy manual work. Besides, the field of artificial intelligence arises as an unavoidable evolution. Computers can intelligently learn, interpret, and process at the level of human thought thanks to artificial intelligence (AI). Deep Learning - Machine Learning is the most rapidly growing AI, and it is inspired by the human brain - the connections between neurons.
Deep Learning has been applied in many real-life problems, breaking the limitations of computers, helping machines to handle or almost exactly like humans. Applying that practice, our team would like to present a report on the application topic of Deep Learning which is "Classification of interior design styles applying deep learning". 9 LIST OF TABLES Table 1: Data statistic. 49 Table 2: Pre-built models provided by Keras.
55 10 LIST OF FIGURES Figure 1: Classification problem. 17 Figure 2:Relatationship between AI, ML and DL. 32 Figure 3: Convolution equation. 33 Figure 4: Examble of Kernel.
34 Figure 5: Convolutional Neural Network. 35 Figure 6: The scope of work automatically exploits the devices in interior design and their design features. 40 Figure 7: A high-level overview of the Early-fusion Blending architecture. 41 Figure 8: Architecture of the team's DeepStyle-Siamese network.
41 Figure 9: An overview of the proposed method. 42 Figure 10: VGG19 architecture. 42 Figure 11: Exception and Inception V3 architectures. 44 Figure 12: Resnet50 architecture.
45 Figure 13: PATH of application. 47 Figure 14: Load data from directory. 48 Figure 15: Python feature. 50 Figure 16: Google colab.
53 Figure 21: Xception Model. 57 Figure 22: Code for train model. 57 Figure 23: Code for plot accuracy. 58 Figure 24: Code for plot loss.
58 Figure 25: Loss and accuracy graph. 59 Figure 26: Confusion matrix. 59 11 Figure 27: System proceed overview. 60 Figure 28: Welcome screen of the application.
61 Figure 29: Main screen of the application. 62 Figure 30:Test API with Postman. 62 Figure 31: Main screen created by ReactJS. 63 Figure 32: Feature introduction screen.
64 Figure 33: Group member introduction screen. 65 Figure 34: Demonstration for application. 66 Figure 35: Evaluate crop image. 67 Figure 36: Client evaluate wrong style 1.
67 Figure 37: Client evaluate wrong style 2. 68 Figure 38: Client evaluate right style. 68 12 Contents SPECIALIZED ESSAY MISSION. 6 CHAPTER 1: DESIGN STYLE CLASSIFICATION PROBLEM .1 House design process .3 Objects and scope .2 Popular design styles.
OVERVIEW ABOUT MACHINE LEARNING, DEEP LEARNING .1 What is Machine Learning? .1 Basic problems in Machine Learning .2 Classify Machine Learning Algorithms .2 What is Deep Learning? .3 Convolutional Neural Network .2 Convolutional Neural Network .3 Basic classes in CNN .4 Structure of CNN. 36 CHAPTER 3: DEEP LEARNING APPLICATION TO SOLVE THE PROBLEM OF CLASSIFICATION OF INTERIOR DESIGN STYLE .1 Research on “Accessing and extracting the characteristics of the interior” .2 Research on “Multimodal search engine for fashion and interior design”. 45 CHAPTER 4: EXPERIMENTS AND RESULTS .1 Reference source for data .3 Description of the dataset .2 Environment and Tools .1 Choose the train model .2 Train the selected model with the dataset .3 Web application development. 68 CHAPTER 5: CONCLUSIONS AND FUTURE WORKS.
70 15 CHAPTER 1: DESIGN STYLE CLASSIFICATION PROBLEM 1.1 House design process 1.1 Research reality People's quality of life is increasingly improving as technology advances. Aside from such advancement, the demand for housing has always been an essential necessity, and it is currently expanding. A house must not only be a location full of facilities to live in, but it must also be beautiful, convenient to use, safe, inexpensive, and affordable. Quality, comprehensive, and intricate design is never easy.
As a result, we investigated the present house design process. The modern house design process typically consists of five essential steps: Customers get advice, make requests, and share ideas with brokers. Brokers establish touch with designers and get consumer needs. The designer receives the request, designs it, and delivers it to the manufacturer.
Manufacturers receive designs, assess feasibility, compile data, and provide estimates to clients. Customers obtain manufacturer information, request adjustments, and approve. The process is iterated until the consumer has reached an agreement with the brokers, designers, and manufacturers.2 Problem The aforementioned method experienced some significant issues as a result of the survey: Idea disagreement. I'm wasting a lot of time.
A huge waste of money. In the exchange process, there is a lack of information.2 Solution The challenge of using information technology to the problems of house design is to: 16 Time has been reduced. Reduce the expenses. Consistency of information Need a solution to help with design and identifying client wants.3 Objects and scope 1.1 Objects The topic focuses on the specific research in the realm of information technology, namely deep learning in analysis to assist customer requirements identification.
Figure 1: Classification problem 1.2 Popular design styles 1.1 What is ArtDecor? The ArtDecor style is an art and decoration school that originated in Paris in the 1920s and extended around the world in the 1930s. ArtDecor strives for basic lines and classic geometric blocks in space, influenced by Cubism, to create a powerful and unique design, similar to German architecture today.2 Characteristics This style's basic background would include: Flowers and leaves with animal skins. Horizontal lines are preferred. Parallelograms, rectangles, and other angle quadrilaterals Image of a stylized animal model.
Sharp geometric angles or representations of stylized architecture Military and zigzag patterns Some are distinguished by achievements in trade, science, and technology.3 Features About color: ArtDecor colors are often vivid and high contrast, such as yellow (bright gold or vintage), red, green, blue, rose silver, black, and other iridescent hues. Furthermore, colors that contrast with the sheen of the wood and lacquered interiors are popular in this design. Concerning the material used: Some materials, such as stainless steel, glass, or animal skin, will be unavoidable in this ArtDecor style. To enhance the luxurious and dazzling beauty, extremely expensive materials such as marble, rare wood, and so on are employed.
18 Concerning the furniture: Not crowded and intricate, but simple, kind, forceful, and charming. The illustration is simple. Details such as materials, lighting, and accessories: Rugs with huge cubes were frequently used to cover hardwood floors. Crystals are frequently used in decorative lights to keep them looking bright and fresh.
Ornamental textiles provide decorative value to an area by using contrasting forms and colors.1 What is HiTech? Hi-tech architecture grew rapidly in the late twentieth and early twenty-first centuries. This movement expanded fast throughout numerous countries. This style thrived in both Japan and the West. Hitech is also known as High Technology.
This design constantly employs modern materials, equipment, furniture, and goods that incorporate cutting-edge technology into space architecture.2 Characteristics Hitech, like many other interior designs, has essential qualities that set it apart from the others: Display aesthetic value by utilizing cutting-edge technologies. The area is contemporary because of the use of simple colors. Simple interior design communicates freedom and refinement. Ensure optimal comfort.3 Features Colors that are simple: Hitech interior design always combines incredibly basic colors with a lot of lines.
Rustic hues are constantly fashionable and used more.