SMOL-MapSeg: Show Me One Label

2025-08-07

AI Summary

Historical maps exhibit heterogeneous visual styles and inconsistent semantic annotations, posing significant challenges for accurate segmentation by existing deep learning models. To address this, we propose SMOL-MapSeg, a segmentation framework centered on an On-Need Declarative (OND) knowledge prompting mechanism: during inference, user-specified concept definitions and morphological priors are dynamically injected—replacing SAM’s original prompt encoder—and the model is fine-tuned on historical map data. This enables zero-shot or few-shot generalization to novel categories without architectural modification, substantially enhancing semantic adaptability. Experiments demonstrate that SMOL-MapSeg outperforms the UNet baseline in mean segmentation accuracy, precisely identifies OND-defined categories, and rapidly generalizes to unseen concepts with only a few annotated samples. The framework thus bridges the gap between domain-specific knowledge and data-efficient segmentation in historical cartography.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multimodal Learning
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Historical maps are valuable for studying changes to the Earth's surface. With the rise of deep learning, models like UNet have been used to extract information from these maps through semantic segmentation. Recently, pre-trained foundation models have shown strong performance across domains such as autonomous driving, medical imaging, and industrial inspection. However, they struggle with historical maps. These models are trained on modern or domain-specific images, where patterns can be tied to predefined concepts through common sense or expert knowledge. Historical maps lack such consistency -- similar concepts can appear in vastly different shapes and styles. To address this, we propose On-Need Declarative (OND) knowledge-based prompting, which introduces explicit prompts to guide the model on what patterns correspond to which concepts. This allows users to specify the target concept and pattern during inference (on-need inference). We implement this by replacing the prompt encoder of the foundation model SAM with our OND prompting mechanism and fine-tune it on historical maps. The resulting model is called SMOL-MapSeg (Show Me One Label). Experiments show that SMOL-MapSeg can accurately segment classes defined by OND knowledge. It can also adapt to unseen classes through few-shot fine-tuning. Additionally, it outperforms a UNet-based baseline in average segmentation performance.

Problem

• Addressing inconsistency in historical map segmentation with deep learning
• Proposing OND prompting to guide concept-pattern recognition
• Improving segmentation accuracy for historical maps over UNet

Innovation

• OND knowledge-based prompting for historical maps
• Replaces SAM prompt encoder with OND mechanism
• Few-shot fine-tuning for unseen classes adaptation