tint_map has negative values resulting in NaN in uvl conversion

[CemEntok]

Hello,
In dataloader_v4.py:

if self.random_color and self.split=='train':
     augment_chroma = self.random_color(illum_count)
     ret_dict["illum_chroma"] *= augment_chroma
     tint_map = mix_chroma(uncalculable_masked_mixmap,augment_chroma,illum_count)
     input_rgb = input_rgb * tint_map    # apply augmentation to input image

random augmentation to input image is applied here. However, when tint_map has negative values, input_rgb image would have also negative values and then:

def rgb2uvl(img_rgb):
    """
    convert 3 channel rgb image into uvl
    """
    epsilon = 1e-8
    img_uvl = np.zeros_like(img_rgb, dtype='float32')
    img_uvl[:,:,2] = np.log(img_rgb[:,:,1] + epsilon)
    img_uvl[:,:,0] = np.log(img_rgb[:,:,0] + epsilon) - img_uvl[:,:,2]   
    img_uvl[:,:,1] = np.log(img_rgb[:,:,2] + epsilon) - img_uvl[:,:,2]

    return img_uvl

image_rgb goes into this function and logarithm gets negative values. Therefore, img_uvl got NaN values.

How would you rain and obtain your results with using augmentation without having NaN issue mentioned here?

A workaround might be limiting tint_map matrix to have 0 for negative elements of itself but this might affect the reproducing your results badly.

Looking forward to hearing from you and thank you!

Cem

[DY112]

Hi, @CemEntok.
Thank you for your interest about our work.
tint_map is a kind of illumination map, which is a combination of illumination chromaticity and mixture map (⍺ in the paper).
Illumination chromaticity is a G=1 normalized vector, and it's R,B component is also positive.
Mixture map, which is a kind of ratio map of intensity channel (we approximate this with G channel) also has only 0 or positive values.
So under normal circumstances, tint_map cannot have negative numbers as values.