DUE: Dynamic Uncertainty-Aware Explanation Supervision via 3D Imputation

arXiv:2403.10831v1 Announce Type: new
Abstract: Explanation supervision aims to enhance deep learning models by integrating additional signals to guide the generation of model explanations, showcasing notable improvements in both the predictability and explainability of the model. However, the application of explanation supervision to higher-dimensional data, such as 3D medical images, remains an under-explored domain. Challenges associated with supervising visual explanations in the presence of an additional dimension include: 1) spatial correlation changed, 2) lack of direct 3D annotations, and 3) …

abstract application arxiv cs.cv data deep learning dynamic explainability guide however images improvements imputation medical supervision type uncertainty via