Cutting-edge vision-language model exceling in retrieving text and metadata from images.
Context-aware chart extraction that can detect 18 classes for chart basic elements, excluding plot elements.
Model for object detection, fine-tuned to detect charts, tables, and titles in documents.
