This is the collation of the Auto-fNIRS toolkit, and the accompanying research report produced during my first year internship at the Physiological Computing and Intelligence lab UCL using on the toolkit. The research report is a short read and useful to understanding the code, but not necessary to using the toolkit.
The Auto-fNIRS toolkit contains a variety of pre-trained Auto-Encoder models for denoising physiological noise from fNIRS signals, but also offers the tools for easy data generation, model training and performance evaluation without requiring skilled knowledge of the subject.