funsd

Here are 4 public repositories matching this topic...

Nimalan07 / nlp-document-intelligence-system

AI-powered document intelligence system using OCR, NLP, Named Entity Recognition, and FastAPI for structured information extraction from scanned documents.

python nlp docker machine-learning ocr computer-vision deep-learning spacy fastapi document-intelligence funsd

Updated Jul 19, 2026
Python

nafis8766 / adaptive-donut-ocr

Star

Learned visual-token pruning for Donut document OCR. 65% cross-attention KV memory cut for -0.26 word-recall points; learned selection beats random by up to +34 points. Pre-registered analysis, executable verifiers, documented negative results.

reproducible-research transformers pytorch donut efficient-inference vision-language ablation-study document-ocr token-pruning funsd

Updated Oct 2, 2026
Jupyter Notebook

Raghavtripathii / docshield-ai

Star

Research-oriented Document AI using LayoutLMv3 + EasyOCR for multimodal document understanding, entity extraction, and robustness evaluation.

python nlp machine-learning ocr computer-vision transformers pytorch information-extraction huggingface document-understanding streamlit model-robustness document-ai token-classification layoutlmv3 multimodal-ai funsd

Updated Aug 9, 2026
Jupyter Notebook

AD2000X / FinDocStructRAG

Star

Layout-aware financial-document intelligence: TATR table extraction, OCR content, DocLayNet layout, FUNSD relations, and table-grounded RAG QA. Honest subset evaluation; artifact-backed Gradio demo.

ocr information-extraction tables layout-analysis rag document-ai table-transformer financial-documents funsd

Updated Jul 10, 2026
Python

Add this topic to your repo

To associate your repository with the funsd topic, visit your repo's landing page and select "manage topics."

Learn more