cs.CVNov 2, 2026

ArabGeoText: Tool-Augmented Vision-Language Models for Arabic Geospatial Text Spotting

Authors: Yakoub Bazi, Sarah Alsaleh, Mohamad M. Al Rahhal, Mohamed A. Mekhtiche, Mansour Zuair

Organizations: Computer Engineering Department, College of Computer and Information Sciences, King Saud University, Riyadh 11543, Saudi Arabia · Applied Computer Science Department, College of Applied Computer Science, King Saud University, Riyadh 11543, Saudi Arabia

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Geospatial text spotting involves locating and transcribing geographic labels, such as road names, district names, and landmarks, in raster maps. These labels are often small, densely arranged, and arbitrarily oriented. Although Arabic text recognition has been studied extensively in documents and natural scenes, Arabic geospatial text spotting remains underexplored. We introduce ARABGEOTEXT, an Arabic-centric benchmark covering 970 geographic locations in Riyadh and Jeddah, with 17,875 manually annotated text instances per view type. For each location, we pair an OpenStreetMap (OSM) rendering with a composite created by overlaying that rendering on the corresponding remote sensing (RS) imagery. The paired views preserve label content and pixel geometry, allowing evaluation across different backgrounds. Geographic partitions support in-city and city-disjoint evaluation of detection, recognition, and end-to-end spotting. Experiments with direct Qwen3.5-4B spotting show difficulties in recovering all labels within a tile. We therefore propose a tool-augmented vision-language framework that combines YOLO26 localization and crop rectification with Qwen3.5-4B recognition adapted through low-rank adaptation (LoRA). The framework outperforms direct LoRA-adapted Qwen3.5-4B spotting across both image views and test cities. With localization held fixed, reader adaptation further improves recognition over the pretrained zero-shot reader. The dataset and code will be made publicly available at https://github.com/yakoubbazi/ArabGeoText.

Figures & tables

Fig. 1: Paired examples from ArabGeoText : (a) OSM renderings and (b) RS+OSM composites. Colored polygons mark annotated text instances. The examples show cropped regions; the displayed instance counts refer to the complete tiles.
Fig. 2: Sample label vocabulary from training, Test-I, and Test-II. Colors indicate Arabic (blue), Latin (orange), mixed-script (magenta), and numeric (gray) labels. Font size does not represent frequency.
Fig. 3: Three paradigms for VLM-based geospatial text spotting: (a) direct spotting, (b) tool-augmented spotting with zero-shot recognition, and (c) tool-augmented spotting with adapted recognition.
Fig. 4: Crop preparation: (a) text region detected by YOLO26, (b) its axis-aligned crop, and (c) the rectified and upscaled crop supplied to the VLM reader.
Fig. 5: Localization examples. Rows: Test-I OSM, Test-I RS+OSM, Test-II OSM, and Test-II RS+OSM. Columns: (a) ground truth, (b) adapted direct Qwen3.5-4B, and (c) YOLO26. Reference polygons are magenta; predictions are blue.
StageQwen3.5-4BQwen3.5-0.8B
YOLO26 localization40.540.5
Crop rectification42.042.1
VLM recognition1619.0917.7
Total per tile1701.51000.3
Recognition per text crop87.449.5
TABLE VII: Average inference time in ms over the test sets on one NVIDIA RTX A6000.
More in this topic

Scene Text Recognition

Scene Text Recognition reads text in natural images and video, including distorted, multilingual, low-resolution, and stylized writing.

See all 29 papers

Explore similar work

CardsList
Jun 5, 2026

Textual Supervision Enhances Geospatial Representations in Vision-Language Models

Geospatial understanding is a critical yet underexplored dimension in the development of machine learning systems for tasks such as image geolocation and spatial reasoning. In this work, we analyze the geospatial representations acquired by three model families: vision-only architectures (e.g., ViT), vision-language models (e.g., CLIP), and large-scale multimodal foundation models (e.g., LLaVA, Qwen, and Gemma). By evaluating across image clusters, including people, landmarks, and everyday objects, grouped based on the degree of localizability, we reveal systematic gaps in spatial accuracy and show that textual supervision enhances the learning of geospatial representations. Our findings suggest the role of language as an effective complementary modality for encoding spatial context and multimodal learning as a key direction for advancing geospatial AI.

Marcelo Sartori Locatelli, Fernando Tonucci, Jea Kwon +5

Max Planck Institute for Security and Privacy (MPI-SP), Bochum, Germany · Universidade Federal de Minas Gerais, Belo Horizonte, Brazil · Changwon National University, Changwon, South Korea +1

Vision-Language ModelsGeospatial Representation LearningMultimodal Large Language ModelsGeospatial ReasoningVLM Reasoning
AskPDF
Mar 4, 2026

TimeSpot: Benchmarking Geo-Temporal Understanding in Vision-Language Models in Real-World Settings

Geo-temporal understanding, the ability to infer location, time, and contextual properties from visual input alone, underpins applications such as disaster management, traffic planning, embodied navigation, world modeling, and geography education. Although recent vision-language models (VLMs) have advanced image geo-localization using cues like landmarks and road signs, their ability to reason about temporal signals and physically grounded spatial cues remains limited. To address this gap, we introduce TimeSpot, a benchmark for evaluating real-world geo-temporal reasoning in VLMs. TimeSpot comprises 1,455 ground-level images from 80 countries and requires structured prediction of temporal attributes (season, month, time of day, daylight phase) and geographic attributes (continent, country, climate zone, environment type, latitude-longitude) directly from visual evidence. It also includes spatial-temporal reasoning tasks that test physical plausibility under real-world uncertainty. Evaluations of state-of-the-art open- and closed-source VLMs show low performance, particularly for temporal inference. While supervised fine-tuning yields improvements, results remain insufficient, highlighting the need for new methods to achieve robust, physically grounded geo-temporal understanding TimeSpot is available at: https://TimeSpot-GT.github.io.

Azmine Toushik Wasi, Shahriyar Zaman Ridoy, Koushik Ahamed Tonmoy +5

Computational Intelligence and Operations Laboratory (CIOL), Bangladesh · Shahjalal University of Science and Technology (SUST), Sylhet, Bangladesh · North South University (NSU), Dhaka, Bangladesh +1

Spatiotemporal ReasoningGeospatial Reasoning BenchmarksGeospatial ReasoningVLM Reasoning
AskPDF
Jul 30, 2026

One Patch Is Enough: Reinforcement-Optimized Visual Token Grounding for MLLM-Based Scene Text Spotting

Scene text spotting requires high-precision alignment between textual recognition and spatial localization. While visual-token grounding has emerged as a promising formulation for Multimodal Large Language Models (MLLMs), the previous multi-patch paradigm often introduces redundant noise and localization ambiguity, particularly for dense or small text instances. To address this, we propose Single-Patch Text Spotting (SPaTS), a vision-centric framework that routes each text instance through a single anchor visual token and then recovers geometry via full-image refinement. To accurately identify this anchor without oracle labels, we introduce Single-Patch Selective Optimization (SPaSO), a reinforcement learning framework that optimizes discrete visual-token selection using patch-level rewards. To further improve representation robustness and localization precision, we introduce Directional Embedding Alignment (DEA) to suppress unstable norm bias by decoupling feature magnitude and direction, and Patch-Enhanced Decoding (PED) to fuse the routed anchor with language semantics and cross-attend over the full-image feature map for geometry-aware boundary regression beyond coordinate-space surrogates. Extensive experiments demonstrate that SPaTS consistently and significantly outperforms both frontier closed-source MLLMs and OCR MLLMs. Code is available at https://github.com/eeNickTang/SPaTS.

Rui Tang, Wentao Yang, Peirong Zhang +4

South China University of Technology · Guangzhou, China · HiThink Research +1

Visual TokenizationScene Text RecognitionLarge Vision-Language ModelsVision-Language GroundingRL for VLMs
AskPDF