Services for processing media files using neural networks are gaining tremendous popularity. Many of them are based on open source libraries.
An example is “Deep-Object-Removal”, which removes the selected object in the photo. This is implemented by restoring the specified area with adjacent colors.
The source code is available at the link:
https://github.com/VPanjeta/Deep-Object-Removal
Installation instructions:
1 . Copy the sources with the command: git clone https://github.com/VPanjeta/Deep-Object-Removal.git
2. Change to the project folder: cd Deep-Object-Removal
3 . Install required libraries: sudo pip3 install -r requirements.txt
four. Run: python main.py
By default, the program requires an image with a size of 400×400, but this value is replaced by the parameter size in the file main.py.
Photos for processing must be placed in the folder “images”. Hot keys for working with the program:
[Esc] – Close the application.
[f]
- Process image.[n]
- Go to next image.[r] – Reset current edits.