Keywords: brain decoding, 3D reconstruction, rotation
Abstract: This paper explores a novel brain decoding model for 3D shape perception through a dual pathway architecture mirroring biological vision. Our bio-inspired approach decomposes 3D visual processing into object identity (ventral pathway) and spatial orientation (dorsal pathway) during continuous rotations. We employ circular regression for angle prediction and develop EEG-conditioned multiview diffusion for 3D reconstruction. Our approach successfully decodes both object identity and spatial orientation from EEG signals and demonstrates feasible 3D reconstruction from neural activity, with interpretability analysis revealing ventral pathway activation patterns that support object recognition performance.
Submission Number: 131
Loading