Hi, thank you for your great work.
The high resolution environment map used in the paper can be trained. Is the final result saved the rgb value of the cube map or the direction vector? Can this environment map be visualized?
self.base = torch.nn.Parameter( 0.5 * torch.ones(6, resolution, resolution, 3, requires_grad=True), )
Hi, thank you for your great work. The high resolution environment map used in the paper can be trained. Is the final result saved the rgb value of the cube map or the direction vector? Can this environment map be visualized? self.base = torch.nn.Parameter( 0.5 * torch.ones(6, resolution, resolution, 3, requires_grad=True), )