PyTorch normalization in Deepstream config

[AastaLLL]

Hi,

The normalization equation used in Deepstream looks like this:
https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_plugin_gst-nvinfer.html

y = net-scale-factor*(x-offsets)

The net-scale-factor is used in the same way as the mean value in pyTorch.
However, we don’t have a configure parameter for std.

Here is a discussion by calculating the corresponding mean and offset value via std.
It’s recommended to check it first:

Image preprocess question

Hi, 1. Assuming uniform distribution for an image(the probability of each value is equal), the std is roughly 74.05 with mean = 127.5. 2. If convert the RGB value to [0,1], the std is near 0.578 with mean = 0.0. v’ = (v-127)/128 3. So for 0.229: | v’ - m’ | ^ 2 ~ (0.578)^2 | v’’ - m’’ | ^2 ~ (0.229)^2 v’’ ≈ v’ / (0.578) * (0.229) v’’ ≈ ( v -127 ) / 128 / 0.578 * 0.229 Mean = 127 net-scale-factor = 1/128/0.578*0.229 = 0.003095 Thanks.

Thanks.