Refining Remote Photoplethysmography Architectures using CKA and Empirical Methods
This work addresses model architecture refinement for rPPG, a domain-specific task in healthcare monitoring, but it is incremental as it applies existing diagnostic methods to a specific application.
The paper tackled the problem of determining optimal model depth in remote photoplethysmography (rPPG) architectures to avoid redundancies and sub-optimal performance, using Centered Kernel Alignment (CKA) to identify that shallower models learn different representations and deeper models add redundant layers, with empirical studies confirming these impacts on performance.
Model architecture refinement is a challenging task in deep learning research fields such as remote photoplethysmography (rPPG). One architectural consideration, the depth of the model, can have significant consequences on the resulting performance. In rPPG models that are overprovisioned with more layers than necessary, redundancies exist, the removal of which can result in faster training and reduced computational load at inference time. With too few layers the models may exhibit sub-optimal error rates. We apply Centered Kernel Alignment (CKA) to an array of rPPG architectures of differing depths, demonstrating that shallower models do not learn the same representations as deeper models, and that after a certain depth, redundant layers are added without significantly increased functionality. An empirical study confirms how the architectural deficiencies discovered using CKA impact performance, and we show how CKA as a diagnostic can be used to refine rPPG architectures.