UpRes ML

Description

The UpRes ML node employs machine learning to resize images while preserving critical features. To train the UpRes model, thousands of images were utilized, resulting in high quality results.

Note: On Linux, the first time you apply UpRes ML, the ML models need to be compiled for your graphics card and resolution being used. The type of card and resolution will determine how long this process will take. Mac users should be running macOS 12.0 or higher for best performance when using the UpRes ML node.

Node Group

AI/ML, Transform.

Controls

Scale

Sets the amount of scaling.

Model

Specifies which machine learning model to use during processing.

Upscale Quality

Sets the quality of the result when using the NVIDIA RTX Super Resolution model.

Sharpen

Sharpens the resized result.

Amount

Determines how much contrast is added at the edges.

Radius

Controls the size of the edges you wish to sharpen.

Threshold

The threshold setting is used to sharpen more pronounced edges, while leaving more subtle edges untouched. Low values sharpen more image areas while higher threshold values sharpen less.