Understanding Proprioception: Neural Network Models

This lecture delves into a paper that explores proprioception, a sense related to self-perception and muscle contraction, using machine learning techniques to understand how the brain functions. The study introduces a new dataset and tasks to explain human proprioception, focusing on trajectory and action recognition tests. Different neural networks are trained and compared to analyze their performance, with a particular emphasis on action recognition models. The lecture concludes that action recognition training models provide insights into how the human brain processes movement and speed, offering a better understanding of proprioception.