Thank you very much for your great work. I find an issue in the function gen_random_rays_at. The purpose of line 160 pixels_y = torch.LongTensor(select_index % (self.W // 8)) * 8 is to generate random integers along the y-axis of values >= 0 and < self.H . But, instead, the statement above obviously generates values in [0, self.W). I think when the case of H != W happens, the random ray generation may go wrong, right?
Hi,
Thank you very much for your great work. I find an issue in the function
gen_random_rays_at. The purpose of line 160pixels_y = torch.LongTensor(select_index % (self.W // 8)) * 8is to generate random integers along the y-axis of values >= 0 and < self.H . But, instead, the statement above obviously generates values in [0, self.W). I think when the case of H != W happens, the random ray generation may go wrong, right?Sorry, I just found your code is correct.