STT | MSSV | Họ và Tên | Github | |
---|---|---|---|---|
1 | 21522634 | Lê Chí Thịnh | lechithinh | [email protected] |
2 | 21522621 | Huỳnh Công Thiện | HuynhThien1 | [email protected] |
3 | 21522706 | Nguyễn Minh Trí | MinhTri17 | [email protected] |
- Tên môn học: Tính toán đa phương tiện
- Mã môn học: CS232
- Mã lớp: CS232.N21.KHCL
- Năm học: HK2 (2022 - 2023)
- Giảng viên: TS.Đỗ Văn Tiến
We are thrilled to introduce our cutting-edge image retrieval system that is powered by advanced algorithms and state-of-the-art machine learning techniques. This remarkable system represents a significant leap forward in the field of image searches, offering users effortless access to highly precise results. It is with great pleasure that we bid farewell to the cumbersome task of endless scrolling and welcome an exciting new era of image discovery.
- Utilize your own dataset for image retrieval.
- Provide a variety of feature descriptors.
- Offer a diverse range of search methods.
- Experiment with your own camera.
- Visualize your results.
Here are the techs we implemented for this project
Clone the repository
git clone https://github.com/lechithinh/Image-Retrival.git
Please organize your code following this structure:
Main-folder/
│
├── Asset/
| | ├── asset1.png
| | ├── asset2.png
| | ├── asset3.png
├── dataset/
| | ├── Black_dress (1).png
| | ├── Black_dress (2).png
| | ├── Black_dress (3).png
| | ├── ....
| | ├── Red_dress(30).png
├── feature/
| | ├── LBP.index.bin
| | ├── LBP.csv
| | ├── RGBHistogram.index.bin
| | ├── RGBHistogram.csv
| | └── VGG16.index.bin
| | └── VGG16.csv
├── app.py
├── ...
├─ main.py
└── ...
The dataset for this project is Fashion Dataset
- There are 24 class
- Each class contains 30 images
- The total images is 720
Be sure follow The same structure
.
- These are features that we extracted based on the dataset provided above Feature
- You you download this one and run your code immediately or you can use the dataset above and extract these features on your own.
- Create the virtual environment
python -m venv venv #activate the virtual environment .\venv\Scripts\activate.bat
- Install required libraries.
pip install -r requirement.txt
- Follow the structure above to configure the dataset.
- Run the web.
streamlit run main.py
- Configure your username and password on
credentials.json
to login.
- Create your dataset.
- Follow the structure to configure the dataset.
- Move into
Extract Features
tab on the website. - Select the
Dataset path
. - Select the
Feature Descriptor
.
- Move into
Search System
. - Select either full image or crop image method.
- Upload your image,
- Configure the
dataset path
- Select the
feature descriptor
- Choose the number of image results.
- Move into
Camera System
. - Select either full image or crop image method.
- Capture your picture.
- Configure the
dataset path
- Select the
feature descriptor
- Choose the number of image results.