Abstract: Vision-language foundation models (VLMs) have shown great potential in feature transfer and generalization across a wide spectrum of medical-related downstream tasks. However, fine-tuning ...
Abstract: Hyperspectral image (HSI) classification in few-shot lifelong learning faces dual challenges: catastrophic forgetting from inaccessible historical training samples and overfitting due to ...
BiLSTM, an ICD-11 automatic coding model using MC-BERT and label attention. Experiments on clinical records show 83.86% ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback