BloodMNIST
PulseAugur coverage of BloodMNIST — every cluster mentioning BloodMNIST across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
New FedSumUp method enhances privacy and efficiency in federated learning
Researchers have developed FedSumUp, a novel method for federated learning that enhances privacy and reduces computational overhead on client devices. FedSumUp utilizes a learnable lossy compressor to generate compact, …
-
MOXIE framework reveals multiple explanations for biomedical image classifiers
Researchers have developed MOXIE, an evolutionary framework designed to uncover multiple alternative explanations for biomedical image classifier predictions. Unlike methods that provide a single explanation based on a …
-
New regularizers improve AI model performance by isolating noise
Researchers have developed new regularizers to improve deep neural network performance by enforcing signal-noise factorization (SNF) and signal-signal factorization (SSF) during training. Experiments on CIFAR-100 showed…
-
StyleGANCA: Lightweight NCA for Medical Image Synthesis
Researchers have introduced StyleGANCA, a novel generative adversarial network that utilizes Neural Cellular Automata (NCA) for medical image synthesis. This architecture integrates a StyleGAN-inspired mapping network w…
-
New GradSkip method enhances Vision Transformer explainability and efficiency
Researchers have developed GradSkip, a new method for explaining the decisions made by Vision Transformers (ViTs). Unlike previous methods that assume uniform importance across attention heads and treat skip connections…
-
DMDSC classifier adapts margins for imbalanced medical image datasets
Researchers have developed a new classifier called DMDSC, designed to improve open-set recognition in medical imaging datasets that suffer from extreme class imbalances. This dynamic-margin approach adjusts margins base…