shuochen365

dists = dists * tf.linalg.norm(rays_d[..., None, :], axis=-1)

5

Hi, Why multiply each distance by the norm of the light in its corresponding direction to convert it into a real-world distance? > dists = dists * tf.linalg.norm(rays_d[..., None, :],...

why sample t0 linearly from 0 to 1 in order to get a linear sampling in disparity

1

I can understand the paper moving the camera position origin to the z = -n plane. > t = -(near + rays_o[..., 2]) / rays_d[..., 2] > rays_o = rays_o...

why sample t0 linearly from 0 to 1 in order to get a linear sampling in disparity

1

I can understand the paper moving the camera position origin to the z = -n plane. > t = -(near + rays_o[..., 2]) / rays_d[..., 2] > rays_o = rays_o...