Q-Mask: Query-driven Causal Masks for Text Anchoring in OCR-Oriented Vision-Language Models
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 limita...
| Authors | Longwei Xu, Feng Feng, Shaojie Zhang et al. |
|---|---|
| Primary category | cs.CV |
| Published | 31 March 2026 |
| Last updated at source | 21 April 2026 |
Connected in the knowledge graph
- loongwayX/TABenchDataset · cited by
Models and datasets referencing this paper
- loongwayX/TABenchDataset
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.