The Ai Denoiser can also be run from the command line on frames that have already been rendered to disc.

$RMANTREE/bin/denoise_batch

usage: denoise run [-h] [-v] [-a ASYMMETRY] [-o OUTPUT] [-cf] [-f] [-j]
                   files [files ...]

positional arguments:
files                
Image files to be denoised. Can accept a glob as input. Ex: foo.*.exr

$RMANTREE/bin/denoise_batch -h
usage: denoise run [-h] [-v] [-a ASYMMETRY] [-o OUTPUT] [-cf] [-f] [-j]
                   files [files ...]

positional arguments:
  files                 Image files to be denoised. Can accept a glob as input. Ex: foo.*.exr

optional arguments:
  -h,   -h,            show this help message and exit
  -v, --verbose         Verbose output for debugging.
  -a ASYMMETRY, --asymmetry ASYMMETRY

                        controls the asymmetry value, 0 is the best quality,
                        higher values encourage the denoiser to avoid
                        over blurring, leading to weaker denoising.
  -o OUTPUT, --output OUTPUT
                        The output directory for the denoised images. The JSON
                        config file(s) will also be written to this directory.
                        If not specified, we write into a subdirectory called
                        "denoised", from the image path.
  -cf, --crossframe     Do crossframe denoising
  -f, --flow            Whether to compute optical flow.
  -j, --json            If provided, use the first input file as the
                        configuration JSON file. All other files and options
                        are ignored.