search

MasterBin-IIAU / Unicorn

[ECCV'22 Oral] Towards Grand Unification of Object Tracking
MIT License
953 stars 87 forks source link

Question regarding input tensor preprocessing #42

Closed jinuhwang closed 1 year ago

jinuhwang commented 1 year ago

Hi,

While following along the inference.py code to see how the model is working, I noticed something in the preprocessing code.

PreprocessorX at external/lib/test/tracker/unicorn_sot.py:111 turns RGB format back to BGR format and the normalization process is missing. The self.normalize is not referenced at all as well. So the input seems to be raw image in BGR format with values between range [0, 255].

I wasn't able to find any code that performs normalization in the forward functions of the inner models as well. Is it just that the model was trained on raw pixel values or am I missing anything?

MasterBin-IIAU commented 1 year ago

Hi, yes, the model takes the original pixel values ranging from 0 to 255 as the inputs. This is aligned with YOLOX, which does not use any normalization.

jinuhwang commented 1 year ago

@MasterBin-IIAU Thank you for the clarification, I really appreciate it.