one or more processors; and

one or more non-transitory computer-readable storage media, coupled with the one or more processors, having instructions stored thereon, which, when executed by the one or more processors, cause the computing system to:

obtain, at an embedded print raster image processor in a printer device, an input associated with an image desired for printing;

identify a region of interest associated with the input based on a particular task to be performed by a lean machine learning model or based on a user selection;

generate, via the lean machine learning model operating in association with the embedded print raster image processor in the printer device, a first output corresponding with the region of interest, wherein the lean machine learning model is trained using a loss value generated in association with training a complex machine learning model having a greater number of layers than the lean machine learning model;

generate, via the embedded print raster image processor in the printer device, a second output corresponding with a remaining portion of the input; and

aggregate the first output corresponding with the region of interest and the second output corresponding with the remaining portion of the image to generate a final output.