Would you please explain how you process the mpii dataset?

[CoinCheung]

Hi,

I am trying to learn your code. Here I do not quite understand what the method of crop in the transforms.py does. Did you transfer the upper left and bottom right point of the original bounding box its rotated and re-scaled version, and then implementing the crop operation?

What does these lines mean please?

    # Upper left point
    ul = np.array(transform([0, 0], center, scale, res, invert=1))
    # Bottom right point
    br = np.array(transform(res, center, scale, res, invert=1))

Why did you use [0,0] and res to be the input of the transform method?