KLA-Denoising-Task

This Repository contains the files for Image Restoration from Noisy and Blur Images using Deep Learning Models

Models Used :

• Denoising Task: RIDNet
• Defect Mask Segmentation: UNet

Model & Dataset Link :

• Dataset: https://drive.google.com/file/d/1iTM57U43L0ANn1u_bRv4dEEERQemuOXc/view?usp=drive_link
• RIDNet: https://drive.google.com/file/d/1cuyH0LxOmmW3_yZsL0GcsdodcwVhfcMy/view?usp=sharing
• UNet: https://drive.google.com/file/d/1pnXopSsV8yAq6GXf0AEsDQII2UdlVU2k/view?usp=sharing

Steps to Organize the Dataset :

• Download the dataset from the link given above

• Use the files.py to organize the dataset into a new folder, dataset with subfolders data, label and mask, where data folder contains the noisy image, label folder contains the clean ground truth image, and mask folder contains the ground truth defect mask

  

• Do change the class name in Line 59 of files.py each time to append the images to dataset folder

Steps to Run the Model :

• Download the required python librariies using requirements.txt

  pip install -r requirements.txt
  

• Clone the Repository

  git clone https://github.com/gokulmk-12/KLA-Denoising-DLI.git
  

• Download the model weights from the above link and paste in a new folder "saved_models" inside the clones folder

• Below is the expected contents of the cloned folder

  

• Run main.py by using the following command. It opens a GUI in streamlit

  streamlit run main.py
  

• Below is a video demonstration on how to use the GUI

  video_demo.mp4

Steps to Modify the Model :

• The main model files used for training are in the models folder
• The user are encouraged to change config.py with thier trained weights, provided they plan to use the same architecture as in networks.py