2 unstable releases
0.3.0 | Mar 27, 2024 |
---|---|
0.2.0 | Aug 29, 2023 |
#173 in Machine learning
19KB
376 lines
Resup
Resup is CLI front-end for realesrgan-ncnn-vulkan
image upscaler written in Rust.
It allows you quickly upscale your images directly in CLI.
Installation
Install Real-ESRGAN
Resup requires realesrgan-ncnn-vulkan
to be installed.
Check your distro repositories for their availability.
If its not available in your distro repositories, download it from official repository.
Install Resup
Install with cargo
Use command below to install it from crates.io (make sure that you have installed Rust toolchain and C/C++ compiler).
cargo install resup
Install from releases
Go to release tab and download version that you want. Extract archive content and place executable in directory that exists in PATH
variable.
Build from source
You can compile Resup manually. Clone this repository, install Rust toolchain and C/C++ compiler and run build with command below.
cargo build --release
Usage
To start upscale you can use upscale
command with the path to your file:
resup upscale my_image.jpg
In this example, image will be saved as my_image-upscaled.png
. You can specify name of output image manually with --output
argument.
resup upscale my_image.jpg --output my_image.png
You can get full list of available models. Use list
command for it:
resup list
Configuration
You can configure Resup for your system without need to edit configuration file. You can use Resup builtin commands for it.
model
- Shows currently used model. To specify new, use this command and as argument provide name of model. Example:resup model realesrgan-x4plus-anime
.models-dir
- Shows current path to directory with models. To specify new, use this command and as argument provide path to directory with models. Example:resup models-dir ~/esrgan/models
.executable
- Shows current path to executable file. To specify new, use this command and as argument provide path to executable. Example:resup executable ~/esrgan/realesrgan-ncnn-vulkan
.
Dependencies
~1.4–8.5MB
~76K SLoC