Neeraj Basu (neerajb)
Zero late days used.
The tetrahedron mesh requires 4 faces and 4 vertices to construct.
The cube mesh requires 12 faces and 8 vertices to construct.
Color 1 = [0.878, 0.407, 0.082].
Color 2 = [0.082, 0.133, 0.878].
R_relative = [[0, 1, 0], [-1, 0, 0], [0, 0, 1]]
T_relative = [0, 0, 0]
Rotate the camera -90 degrees about the z-axis.
R_relative = [[1, 0, 0], [0, 1, 0], [0, 0, 1]]
T_relative = [0, 0, 2]
Translate the camera 2 units along the z-axis.
R_relative = [[1, 0, 0], [0, 1, 0], [0, 0, 1]]
T_relative = [0.5, -0.5, 0]
Translate the camera 0.5 units along the x-axis and -0.5 units along the y-axis.
R_relative = [[0, 0, 1], [0, 1, 0], [-1, 0, 0]]
T_relative = [-3, 0, 3]
Rotate the camera -90 degrees about the y-axis then translate -3 units along the x-axis and 3 units along the z-axis.
Meshes contain information about connectivity and faces unlike pointclouds.
Points clouds can be easier to render because there is no texture information associated with each point.
It's easier to deform pointclouds from shape to shape since there is no connectivity information.
Holes in meshes can be unforgiving and cause dramatic spikes on the sufaces of objects.
Generally speaking, pointclouds are easier to acquire than meshes.
I call this piece "man who stood up too fast".
Cow pointcloud with 10 sampled points.
Cow pointcloud with 100 sampled points.
Cow pointcloud with 1000 sampled points.
Cow pointcloud with 10000 sampled points.