Confused about convolutional layer implementation in YOLOv3

[s-ilic]

Hi all,

I'm a bit confused about the implementation of the convolutional function in 4-Object_Detection/YOLOV3/core/common.py : how come only the first and last element of the input list filters_shape are used, despite the list having a priori 4 elements? Is that a typo ?

Thanks in advance !