MedMNIST
PulseAugur coverage of MedMNIST — every cluster mentioning MedMNIST across labs, papers, and developer communities, ranked by signal.
-
Pandora's Regret: A Proper Scoring Rule for Evaluating Sequential Search
Researchers have introduced Pandora's Regret, a novel scoring rule designed to evaluate sequential search processes more effectively than traditional methods. Unlike local rules like log loss, Pandora's Regret considers…
-
Deep neural networks combine Fisher Vectors with CNNs and ViTs for medical image classification
Researchers have developed a novel approach to enhance deep neural networks for medical image classification by integrating Fisher Vectors with hybrid CNN-ViT architectures. This method aims to improve performance on da…
-
UniMo framework uses deep learning for unified medical image motion correction
Researchers have developed UniMo, a novel deep learning framework designed to correct motion artifacts in medical imaging. This unified approach combines an equivariant neural network for global rigid motion and an enco…
-
研究人员提出使用模糊逻辑通过知识发现实现鲁棒图像识别
研究人员开发了一种新颖的方法，通过将领域知识集成到深度神经网络中来增强图像识别的鲁棒性。该方法引入了一个可微分知识单元（DKU），它使用模糊逻辑和蕴含规则来调制分类器的logits，以优化类概率。该系统能够从任务监督中自动发现隐式概念，从而在不需要显式概念标签的情况下学习类与这些概念之间的关系。在PASCAL-VOC、COCO和MedMNIST数据集上的评估表明，该方法在性能和领域泛化能力方面均有所提高。
-
New AI training method achieves error-free classification on medical datasets
Researchers have developed a novel method called Artificial Special Intelligence (ASI) to train machine learning models for classification tasks without errors. This approach aims to prevent models from repeating mistak…
-
Biomedical AI models learn nonrobust features, impacting accuracy and robustness trade-offs
A new study published on arXiv investigates the presence and impact of nonrobust features in deep learning models used for biomedical image analysis. The research indicates that these nonrobust features, which are predi…