Q-Mask: Query-driven Causal Masks for Text Anchoring in OCR-Oriented Vision-Language Models

• URL: http://arxiv.org/abs/2604.00161v1
• Date: Tue, 31 Mar 2026 19:09:55 GMT
• Title: Q-Mask: Query-driven Causal Masks for Text Anchoring in OCR-Oriented Vision-Language Models
• Abstract summary: We propose a precise OCR framework built upon a causal query-driven mask decoder (CQMD)<n>Q-Mask performs causal visual decoding that sequentially generates query-conditioned visual masks before producing the final OCR output.<n>To train CQMD, we construct TextAnchor-26M, a large-scale dataset of image-text pairs annotated with fine-grained masks corresponding to specific textual elements.
• Score: 20.508411710998313
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Optical Character Recognition (OCR) is increasingly regarded as a foundational capability for modern vision-language models (VLMs), enabling them not only to read text in images but also to support downstream reasoning in real-world visual question answering (VQA). However, practical applications further require reliable text anchors, i.e., accurately grounding queried text to its corresponding spatial region. To systematically evaluate this capability, we introduce TextAnchor-Bench (TABench), a benchmark for fine-grained text-region grounding, which reveals that both general-purpose and OCR-specific VLMs still struggle to establish accurate and stable text anchors. To address this limitation, we propose Q-Mask, a precise OCR framework built upon a causal query-driven mask decoder (CQMD). Inspired by chain-of-thought reasoning, Q-Mask performs causal visual decoding that sequentially generates query-conditioned visual masks before producing the final OCR output. This visual CoT paradigm disentangles where the text is from what the text is, enforcing grounded evidence acquisition prior to recognition and enabling explicit text anchor construction during inference. To train CQMD, we construct TextAnchor-26M, a large-scale dataset of image-text pairs annotated with fine-grained masks corresponding to specific textual elements, encouraging stable text-region correspondences and injecting strong spatial priors into VLM training. Extensive experiments demonstrate that Q-Mask substantially improves text anchoring and understanding across diverse visual scenes.

Related papers

• An Effective Data Augmentation Method by Asking Questions about Scene Text Images [5.189562992500781]
  We propose a VQA-inspired data augmentation framework that strengthens OCR training through structured question-answering tasks.<n>For each image-text pair, we generate natural-language questions probing character-level attributes such as presence, position, and frequency.<n>These auxiliary tasks encourage finer-grained reasoning, and the OCR model aligns visual features with textual queries to jointly reason over images and questions.
  arXiv  Detail & Related papers  (2026-03-03T23:18:53Z)
• VISTA-Bench: Do Vision-Language Models Really Understand Visualized Text as Well as Pure Text? [51.02924254085878]
  Vision-Language Models (VLMs) have achieved impressive performance in cross-modal understanding across textual and visual inputs.<n>We introduce VISTA-Bench, a benchmark from multimodal perception, reasoning, to unimodal understanding domains.
  arXiv  Detail & Related papers  (2026-02-04T17:48:55Z)
• When Text-as-Vision Meets Semantic IDs in Generative Recommendation: An Empirical Study [48.67151986743594]
  We revisit representation design for Semantic ID learning by treating text as a visual signal.<n>We conduct a systematic empirical study of OCR-based text representations, obtained by rendering item descriptions into images.<n>We find that OCR-text consistently matches or surpasses standard text embeddings for Semantic ID learning in both unimodal and multimodal settings.
  arXiv  Detail & Related papers  (2026-01-21T06:18:57Z)
• Text4Seg++: Advancing Image Segmentation via Generative Language Modeling [52.07442359419673]
  We propose a novel text-as-mask paradigm that casts image segmentation as a text generation problem.<n>Key innovation is semantic descriptors, a new textual representation of segmentation masks.<n>Experiments on natural and remote sensing datasets show that Text4Seg++ consistently outperforms state-of-the-art models.
  arXiv  Detail & Related papers  (2025-09-08T04:07:14Z)
• Knowing Where to Focus: Attention-Guided Alignment for Text-based Person Search [64.15205542003056]
  We introduce Attention-Guided Alignment (AGA) framework featuring two innovative components: Attention-Guided Mask (AGM) Modeling and Text Enrichment Module (TEM)<n>AGA achieves new state-of-the-art results with Rank-1 accuracy reaching 78.36%, 67.31%, and 67.4% on CUHK-PEDES, ICFG-PEDES, and RSTP, respectively.
  arXiv  Detail & Related papers  (2024-12-19T17:51:49Z)
• See then Tell: Enhancing Key Information Extraction with Vision Grounding [32.445618057103324]
  We introduce STNet (See then Tell Net), a novel end-to-end model designed to deliver precise answers with relevant vision grounding.<n>To enhance the model's seeing capabilities, we collect extensive structured table recognition datasets.<n>Our approach demonstrates substantial advancements in KIE performance, achieving state-of-the-art results on publicly available datasets.
  arXiv  Detail & Related papers  (2024-09-29T06:21:05Z)
• Decoder Pre-Training with only Text for Scene Text Recognition [54.93037783663204]
  Scene text recognition (STR) pre-training methods have achieved remarkable progress, primarily relying on synthetic datasets. We introduce a novel method named Decoder Pre-training with only text for STR (DPTR) DPTR treats text embeddings produced by the CLIP text encoder as pseudo visual embeddings and uses them to pre-train the decoder.
  arXiv  Detail & Related papers  (2024-08-11T06:36:42Z)
• Class-Aware Mask-Guided Feature Refinement for Scene Text Recognition [56.968108142307976]
  We propose a novel approach called Class-Aware Mask-guided feature refinement (CAM) Our approach introduces canonical class-aware glyph masks to suppress background and text style noise. By enhancing the alignment between the canonical mask feature and the text feature, the module ensures more effective fusion.
  arXiv  Detail & Related papers  (2024-02-21T09:22:45Z)
• Look, Read and Ask: Learning to Ask Questions by Reading Text in Images [3.3972119795940525]
  We present a novel problem of text-based visual question generation or TextVQG. To address TextVQG, we present an OCR consistent visual question generation model that Looks into the visual content, Reads the scene text, and Asks a relevant and meaningful natural language question.
  arXiv  Detail & Related papers  (2022-11-23T13:52:46Z)
• TextOCR: Towards large-scale end-to-end reasoning for arbitrary-shaped scene text [23.04601165885908]
  We propose TextOCR, an arbitrary-shaped scene text detection and recognition with 900k annotated words collected on real images. We show that current state-of-the-art text-recognition (OCR) models fail to perform well on TextOCR. We use a TextOCR trained OCR model to create PixelM4C model which can do scene text based reasoning on an image in an end-to-end fashion.
  arXiv  Detail & Related papers  (2021-05-12T07:50:42Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.