Select Data to Process

Select additional images to denoise using your trained model.

When processing additional images, the trained model applies what it learned about noise patterns to new data.

File Requirements

The file is validated the same way as the training input:

• Format: Multi-page TIFF (.tif, .tiff)

• Bit Depth: 8-bit or 16-bit grayscale. Float32/float64 data is accepted with a warning and normalized before processing.

• Minimum size: at least 64×64 pixels per slice

• Minimum slices: at least 10 slices

Within those limits the image is processed in patches, so larger stacks are fine.

Critical Consideration

The model works best on images with the same noise characteristics as your training data. For optimal results:

• Use images from the same acquisition session

• Same microscope/camera settings

• Same sample preparation method

• Similar imaging conditions

If your new images have different noise patterns (different detector, different settings, etc.), the model may not perform as well. In that case, consider training a new model on the new data.

What NOT to expect

A model trained on one type of noise won't effectively remove a different type of noise. Each noise source has unique characteristics the model must learn.