Computational model for human 3D shape perception from a single specular image
Takeaki Shimokawa, Akiko Nishio, Masa-aki Sato, Mitsuo Kawato, Hidehiko Komatsu
This study develops a computational model to estimate 3D shapes of specular objects from a single image, showing high accuracy and similarity to human perception errors.
Frontiers in Computational Neuroscience · Mar 1, 2019Peer-reviewedNeuroscience