Grounding Multimodal Evidence
Connect visual claims to source pages and regions, and recognize what an image cannot establish.
Huang Tzu Lin
CategoryBlog
Document & Multimodal Intelligence
Beyond OCR: layout, tables, vision-language models, figure grounding, and cross-modal evidence.
2 posts
Reading Documents Beyond OCR
Reconstruct reading order, tables, and source relationships so a document becomes traceable evidence.
Huang Tzu Lin