Learning 3D Representations From 2D Pre-Trained Models via Image-to-Point Masked Autoencoders

Devs

Learning 3D Representations From 2D Pre-Trained Models via Image-to-Point Masked Autoencoders | Read Paper on Bytez