FeatureZ preserves a broad class of geometric and topological features during lossy compression by enforcing pointwise bounds and consistent star classifications, including features no previous compressor targeted.
Title resolution pending
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 2
citation-polarity summary
roles
background 2representative citing papers
A composite filter ensemble with adaptive profile selection achieved the best balanced accuracy, macro-F1, and Cohen's kappa on all five tested molecular classification datasets, but only against six basic filter baselines.
RespGeomLib generates analysis-ready airway lumen geometry via parametric port assembly and implicit smooth-min blending, producing cleaner junctions and lower memory use than Boolean or global implicit baselines.
A detachable piezoelectric clip on a face mask reads speech vibrations through the mask material, achieving 6.1% character error rate in noise versus 19.7% for a conventional pin microphone.
MERA, a self-supervised Vision Transformer pipeline with sparse-seeded active learning, reports 86.2% malignancy accuracy on LIDC with roughly 1% of training labels, close to its 87.6% fully supervised accuracy, while producing global, case-based, visual, and concept explanations.
An nnU-Net model trained only on public CT data matches human-level kidney and abnormality segmentation and beats TotalSegmentator and BAMF on most tested datasets.
A literature survey organizes healthcare uncertainty quantification methods by ML pipeline stage and finds most published work concentrates on training-stage medical imaging.
citing papers explorer
-
FeatureZ: A General Framework for Feature-Preserving Compression via Pointwise Bounds and Star Classification
FeatureZ preserves a broad class of geometric and topological features during lossy compression by enforcing pointwise bounds and consistent star classifications, including features no previous compressor targeted.
-
ARISE: An adaptive residual-informed stability ensemble for feature selection in small-sample biomedical omics
A composite filter ensemble with adaptive profile selection achieved the best balanced accuracy, macro-F1, and Cohen's kappa on all five tested molecular classification datasets, but only against six basic filter baselines.
-
RespGeomLib: A Reproducible Parametric Engine for Generating Analysis-Ready Human Airway Lumen Geometry
RespGeomLib generates analysis-ready airway lumen geometry via parametric port assembly and implicit smooth-min blending, producing cleaner junctions and lower memory use than Boolean or global implicit baselines.
-
MaskClip: Detachable Clip-on Piezoelectric Sensing of Mask Surface Vibrations for Real-time Noise-Robust Speech Input
A detachable piezoelectric clip on a face mask reads speech vibrations through the mask material, achieving 6.1% character error rate in noise versus 19.7% for a conventional pin microphone.
-
MERA: Multimodal and Multiscale Self-Explanatory Model with Considerably Reduced Annotation for Lung Nodule Diagnosis
MERA, a self-supervised Vision Transformer pipeline with sparse-seeded active learning, reports 86.2% malignancy accuracy on LIDC with roughly 1% of training labels, close to its 87.6% fully supervised accuracy, while producing global, case-based, visual, and concept explanations.
-
Robust Renal Mass Segmentation on CT: A Validation Study of an AI-Based Framework
An nnU-Net model trained only on public CT data matches human-level kidney and abnormality segmentation and beats TotalSegmentator and BAMF on most tested datasets.
-
Uncertainty Quantification for Machine Learning in Healthcare: A Survey
A literature survey organizes healthcare uncertainty quantification methods by ML pipeline stage and finds most published work concentrates on training-stage medical imaging.
- Unifying biomedical knowledge in a modern multimodal graph