FRIEDA: Benchmarking Multi-Step Cartographic Reasoning in Vision-Language Models

Cartographic reasoning is the skill of interpreting geographic relationships by aligning legends, map scales, compass directions, map texts, and geometries across one or more map images. Although essential as a concrete cognitive capability and for critical tasks such as disaster response and urban planning, it remains largely unevaluated. Building on progress in chart and infographic understanding, recent large vision language model studies on map visual question-answering often treat maps as a special case of charts. In contrast, map VQA demands comprehension of layered symbology (e.g., symbols, geometries, and text labels) as well as spatial relations tied to orientation and distance that often span multiple maps and are not captured by chart-style evaluations. To address this gap, we introduce FRIEDA, a benchmark for testing complex open-ended cartographic reasoning in LVLMs. FRIEDA sources real map images from documents and reports in various domains and geographical areas. Following classifications in Geographic Information System (GIS) literature, FRIEDA targets all three categories of spatial relations: topological (border, equal, intersect, within), metric (distance), and directional (orientation). All questions require multi-step inference, and many require cross-map grounding and reasoning. We evaluate eleven state-of-the-art LVLMs under two settings: (1) the direct setting, where we provide the maps relevant to the question, and (2) the contextual setting, where the model may have to identify the maps relevant to the question before reasoning. Even the strongest models, Gemini-2.5-Pro and GPT-5-Think, achieve only 38.20% and 37.20% accuracy, respectively, far below human performance of 84.87%. These results reveal a persistent gap in multi-step cartographic reasoning, positioning FRIEDA as a rigorous benchmark to drive progress on spatial intelligence in LVLMs.

StreetLens: Enabling Human-Centered AI Agents for Neighborhood Assessment from Street View Imagery

Traditionally, neighborhood studies have employed interviews, surveys, and manual image annotation guided by detailed protocols to identify environmental characteristics, including physical disorder, decay, street safety, and sociocultural symbols, and to examine their impact on developmental and health outcomes. While these methods yield rich insights, they are time-consuming and require intensive expert intervention. Recent technological advances, including vision-language models (VLMs), have begun to automate parts of this process; however, existing efforts are often ad hoc and lack adaptability across research designs and geographic contexts. In this demo paper, we present StreetLens, a human-centered, researcher-configurable workflow that embeds relevant social science expertise in a VLM for scalable neighborhood environmental assessments. StreetLens mimics the process of trained human coders by grounding the analysis in questions derived from established interview protocols, retrieving relevant street view imagery (SVI), and generating a wide spectrum of semantic annotations from objective features (e.g., the number of cars) to subjective perceptions (e.g., the sense of disorder in an image). By enabling researchers to define the VLM's role through domain-informed prompting, StreetLens places domain knowledge at the core of the analysis process. It also supports the integration of prior survey data to enhance robustness and expand the range of characteristics assessed across diverse settings. We provide a Google Colab notebook to make StreetLens accessible and extensible for researchers working with public or custom SVI datasets. StreetLens represents a shift toward flexible, agentic AI systems that work closely with researchers to accelerate and scale neighborhood studies.

The mapKurator System: A Complete Pipeline for Extracting and Linking Text from Historical Maps

Scanned historical maps in libraries and archives are valuable repositories of geographic data that often do not exist elsewhere. Despite the potential of machine learning tools like the Google Vision APIs for automatically transcribing text from these maps into machine-readable formats, they do not work well with large-sized images (e.g., high-resolution scanned documents), cannot infer the relation between the recognized text and other datasets, and are challenging to integrate with post-processing tools. This paper introduces the mapKurator system, an end-to-end system integrating machine learning models with a comprehensive data processing pipeline. mapKurator empowers automated extraction, post-processing, and linkage of text labels from large numbers of large-dimension historical map scans. The output data, comprising bounding polygons and recognized text, is in the standard GeoJSON format, making it easily modifiable within Geographic Information Systems (GIS). The proposed system allows users to quickly generate valuable data from large numbers of historical maps for in-depth analysis of the map content and, in turn, encourages map findability, accessibility, interoperability, and reusability (FAIR principles). We deployed the mapKurator system and enabled the processing of over 60,000 maps and over 100 million text/place names in the David Rumsey Historical Map collection. We also demonstrated a seamless integration of mapKurator with a collaborative web platform to enable accessing automated approaches for extracting and linking text labels from historical map scans and collective work to improve the results.