deep-text-recognition-benchmark
deep-text-recognition-benchmark copied to clipboard
Questions about Delta_c
Thank you for your excellent thesis.
One question arose while referring to the code for the paper.
In the original text, I would like you to answer why the order of delta_C required for Transformation calculation and the order of codes are different. Isn't it correct that the order of the 2nd and 3rd np.concatenate should be changed in the code below?
delta_C = np.concatenate( # F+3 x F+3 [ np.concatenate([np.ones((F, 1)), C, hat_C], axis=1), # F x F+3 np.concatenate([np.zeros((2, 3)), np.transpose(C)], axis=1), # 2 x F+3 np.concatenate([np.zeros((1, 3)), np.ones((1, F))], axis=1) # 1 x F+3 ], axis=0 )