#vision-language-model

[ follow ]
#echocardiography #contrastive-learning #multi-view-imaging #document-understanding #layout-preservation
fromwww.nature.com
1 week ago
Medicine

Comprehensive echocardiogram evaluation with view primed vision language AI

EchoPrime is a multi-view, video-based vision-language foundation model trained on over 12 million echocardiogram video-report pairs to enable holistic, view-informed clinical interpretation.
Tech industry
fromInfoQ
1 month ago

IBM Releases Granite-Docling-258M, a Compact Vision-Language Model for Precise Document Conversion

Granite-Docling-258M is an open-source 258M-parameter vision-language model that parses documents with high fidelity while preserving complex layouts, tables, equations, and lists.
[ Load more ]