Skip to main content
Search

# Optical Character Recognition

3 related articles

Subscribe to updates
  1. arXiv

    SP-DocReader cuts Qwen3-VL-4B Vary-600K character error rate from 2.40 to 1.11 with difference-aware self-play

    SP-DocReader is a self-play framework for document OCR that aligns reference and generated readings with a longest common subsequence, applies contrastive relative scoring and direct negative log-likelihood supervision only at unmatched positions, and trains only the OCR module while the visual encoder and language backbone stay frozen; across Qwen3-VL-4B and InternVL-3.5-4B, three self-play rounds reduce Vary-600K character error rate relative to two-epoch supervised fine-tuning and improve DocBank, IIT-CDIP, DocVQA and InfographicVQA scores.

  2. Universal Doc AI: A Universal Intelligent Document Management System Combining OCR, RAG, and LLMs

    This paper presents Universal Doc AI, a multimodal intelligent document management system that unifies Optical Character Recognition (OCR), Retrieval-Augmented Generation (RAG), vector-based semantic search, and Large Language Models (LLMs) into a single platform that extracts text from user-uploaded PDFs, scans, images, and handwritten notes, chunks and embeds it for vector indexing, and answers natural-language questions grounded in the uploaded documents, reporting improvements in retrieval accuracy, response relevance, and interaction efficiency over traditional keyword-based search.

  3. arXiv

    Using OCR Heads to Verbalize Image Semantics: General Verbalization Heads and a Verbalization Lens in VLMs

    Across four vision-language models (Qwen3-VL-2B, Qwen3-VL-8B, Molmo2-7B, Llava-Next-34B), the work identifies attention heads causally necessary for OCR and shows they are general-purpose verbalization heads; summing their output-value matrices yields a verbalization lens that decodes image-token hidden states into interpretable semantic words at any layer, including layer 0, and whose pseudo-inverse provides concept vectors for editing image representations (e.g., replacing a tractor with a revolver), supporting the view that image representations are aligned with language space from early layers.

You’re all caught up