This repository is a replication-focused adaptation of the EMNLP 2023 model for multimodal aphasia type detection, configured for a custom AphasiaBank-derived corpus.
A robust Python toolkit for converting femur segmentation masks (NIfTI format) into high-quality, watertight 3D meshes with automatic cropping, quality filtering, and left/right separation.
Abstract: Brain-to-text systems are rapidly emerging at the intersection of neuroscience, brain–computer interfaces (BCIs), and artificial intelligence, aiming to decode neural signals into coherent ...