Color Perception¶

The utility function we will review is odak.learn.perception.color_conversion.primaries_to_lms():

Let us use this utility function to show how we can obtain LMS sensation from the color primaries of an image:

import odak # (1)
import torch
import sys
from odak.learn.perception.color_conversion import display_color_hvs
from odak.learn.tools import load_image, save_image, resize 

num_primaries = 3
image_primaries = torch.rand(1,
                             num_primaries,
                             8,
                             8
                             ) # (2)

multi_spectrum = torch.rand(num_primaries,
                            301
                           ) # (3)
device_ = torch.device('cpu') # (4)
display_color = display_color_hvs(read_spectrum ='tensor',
                                  primaries_spectrum=multi_spectrum,
                                  device = device_)
lms_color = display_color.primaries_to_lms(image_primaries)  # (5)

Adding odak to our imports.
Generating arbitrary target primaries (the sample image).
Generate arbitrary primaries spectrum
Select your device where tensors will be allocated
Obtain LMS cone sensation using odak.learn.perception.color_conversion.primaries_to_lms

Image title — Sample Generated Image Primaries