Image Inpainting Demo

Web Framework - Jayleen Zhou

Inpainting Algorithm - Lynn Duan, Lei Pan, Jingwen Zhang

---

sitemap inside the working folder

├── README.md
├── app.py
├── server/
    ├── torch_deform_conv/
    ├── predict.py
    ├── <model_file>
    ├── models.py
    ├── layers.py
    └── utils.html
└── static/
    ├── dist/
    ├── script/
        ├── index.jsx
        └── main.jsx
    ├── style/
        └── main.css
    └── index.html

---

How to run this demo locally

• in the project root folder, run python app.py in Terminal
• url: localhost:5000

How to test front-end framework

• change directory into the static folder, run npm install and then npm run dev in Terminal
• url: localhost:8080
• remember to clear browser cache

How to compile front-end

• change directory into the static folder, run npm run build in Terminal

---

Approaches

• Front-end: using ReactJS to render the front-end buttons, jQuery to handle AJAX communications
• Back-end: using Python in Flask framework to serve as the back-end server.

---

Acknowledgement

• Original code of Generator and Discriminators: https://github.com/otenim/GLCIC-PyTorch
• Original code of Deformable Convolutional layer: https://github.com/oeway/pytorch-deform-conv