/SETransformerFusion-A-Multimodal-Fusion-Model-for-Image-and-Blood-Data

SETransformerFusion is a novel multimodal fusion model that combines visual features from images and blood test data to perform classification tasks. It leverages the power of Squeeze-and-Excitation (SE) blocks and Transformer architecture to effectively capture and fuse the information from both modalities.

Primary LanguagePythonMIT LicenseMIT

Watchers