3 ms·
Next step could be recognizing and generalizing shapes (human, car, house), deformation/movement models (rigging, inverse kinematics) and decompositing textures
by rasz 6y ago
Next step could be recognizing and generalizing shapes (human, car, house), deformation/movement models (rigging, inverse kinematics) and decompositing textures into layers (skin, clothes). You would keep common shape/movement/texture atlas in the base codec. As you wrote, the process of recording as a reverse game engine.