A self-supervised approach uses consistent spatial relationships of anatomical structures across patients to improve 3D multi-modal medical image representations, yielding modest gains on segmentation and classification tasks.
Densedino: boosting dense self-supervised learning with token-based point-level consistency
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2verdicts
UNVERDICTED 2representative citing papers
A Marr-inspired boundary-and-surface pretraining stage before contrastive learning yields 2x faster convergence on ResNet18 plus better downstream representations and robustness.
citing papers explorer
-
Beyond Instance-Level Self-Supervision in 3D Multi-Modal Medical Imaging
A self-supervised approach uses consistent spatial relationships of anatomical structures across patients to improve 3D multi-modal medical image representations, yielding modest gains on segmentation and classification tasks.
-
Perceptual Inductive Bias Is What You Need Before Contrastive Learning
A Marr-inspired boundary-and-surface pretraining stage before contrastive learning yields 2x faster convergence on ResNet18 plus better downstream representations and robustness.