fNIRS-Vise is an ambitious research project that pushes the boundaries of brain-computer interfaces (BCIs) by decoding and reconstructing visual experiences directly from fNIRS (functional near-infrared spectroscopy) brain signals.
๐ฏ Key Objectives Decode Visual Stimuli: Develop and refine deep learning models (transformers, autoencoders) to accurately interpret fNIRS data and extract meaningful visual representations. Reconstruct Visual Experiences: Utilize state-of-the-art generative models like Stable Diffusion to transform decoded neural patterns into visual imagery, effectively "seeing" through the mind's eye. Advance fNIRS Applications: Contribute to the growing field of fNIRS research by demonstrating the feasibility of visual decoding and paving the way for novel BCI applications.
๐ง Data & Methodology Proprietary fNIRS Data: High-quality fNIRS recordings collected during visual stimulation tasks, providing a unique resource for model training and validation. Open-Source fNIRS Datasets: Leveraging publicly available datasets (fNIRS2MW, etc.) to enhance model generalizability and robustness. Hybrid Deep Learning Architecture: Integrating the strengths of fNIRS-T, fNIRSNet, and MinD-Vis into a novel architecture optimized for fNIRS vision decoding. Transfer Learning: Utilizing knowledge from fMRI-based visual decoding (MinD-Vis) to accelerate model development and improve performance.
๐ก Inspired By MinD-Vis: A groundbreaking fMRI-based visual reconstruction framework that serves as a key inspiration and reference point for NIRS-Vis.
๐ Broader Impact fNIRS-Vise has the potential to revolutionize our understanding of the human visual system and unlock new possibilities in:
Brain-Computer Interfaces: Enabling more intuitive and immersive communication and control systems. Neurological Research: Providing insights into the neural mechanisms of visual perception and disorders. Clinical Applications: Developing diagnostic and therapeutic tools for visual impairments.
๐ค Get Involved We welcome collaborations and contributions from researchers, developers, and enthusiasts passionate about brain decoding and BCIs. Contact us to explore potential partnerships!
๐ License This project is licensed under the Apache License, Version 2.0. See the LICENSE file for details.