MapSR: Prompt-Driven Land Cover Map Super-Resolution via Vision Foundation Models

2026-04-15

AI Summary

High-resolution land cover mapping is hindered by the prohibitive cost of pixel-level annotations. This work proposes MapSR, a novel framework that introduces prompt-based mechanisms to map super-resolution for the first time. By leveraging a frozen vision foundation model, MapSR extracts class-specific prompts from low-resolution labels and combines linear probing, cosine similarity matching, and graph propagation to enable training-free inference without high-resolution supervision. The approach decouples supervisory signals from model training, reducing trainable parameters by four orders of magnitude and shrinking training time from hours to minutes. Evaluated on the Chesapeake Bay dataset, MapSR achieves a mean Intersection-over-Union (mIoU) of 59.64%, demonstrating substantial gains in efficiency and scalability while maintaining competitive accuracy.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Semi-Supervised Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

High-resolution (HR) land-cover mapping is often constrained by the high cost of dense HR annotations. We revisit this problem from the perspective of map super-resolution, which enhances coarse low-resolution (LR) land-cover products into HR maps at the resolution of the input imagery. Existing weakly supervised methods can leverage LR labels, but they typically use them to retrain dense predictors with substantial computational cost. We propose MapSR, a prompt-driven framework that decouples supervision from model training. MapSR uses LR labels once to extract class prompts from frozen vision foundation model features through a lightweight linear probe, after which HR mapping proceeds via training-free metric inference and graph-based prediction refinement. Specifically, class prompts are estimated by aggregating high-confidence HR features identified by the linear probe, and HR predictions are obtained by cosine-similarity matching followed by graph-based propagation for spatial refinement. Experiments on the Chesapeake Bay dataset show that MapSR achieves 59.64% mIoU without any HR labels, remaining competitive with the strongest weakly supervised baseline and surpassing a fully supervised baseline. Notably, MapSR reduces trainable parameters by four orders of magnitude and shortens training time from hours to minutes, enabling scalable HR mapping under limited annotation and compute budgets. The code is available at https://github.com/rikirikirikiriki/MapSR.

Problem

• land cover mapping
• super-resolution
• weak supervision
• high-resolution
• annotation scarcity

Innovation

• prompt-driven
• vision foundation model
• map super-resolution
• weakly supervised learning
• training-free inference