Quantum inspired networks improve high resolution land cover maps
Quantum-Inspired Convolutional Networks for High-Resolution Pseudo-Label Generation from Multimodal Earth Observation Data
Machine LearningComputer Vision and Pattern Recognition
Summary
Getting detailed maps of land types from satellite images is hard when detailed labels are missing. The authors created a special kind of computer network inspired by quantum computing ideas that combines different kinds of satellite data to guess detailed land cover maps from rough, low-resolution ones. Their method first learns about the input images on its own, then uses rough maps to create more accurate, detailed label maps. Tests showed their approach makes clearer, more consistent land maps than older methods.
What this means in practice
- •For remote sensing analysts: Generate higher-resolution land-cover maps from coarse annotated data and multiple satellite sensors to improve environmental monitoring accuracy.
- •For urban planning teams: Obtain detailed land-use data from blended satellite sources to support planning decisions in areas lacking detailed local maps.
Authors
Abhishek Singh, Giovanni Cuozzo, Alexander Jacob, Claudia Notarnicola
Abstract
Accurate land-cover mapping from multimodal Earth observation (EO) data remains a challenge in regions where high-resolution annotations are scarce or costly to obtain. This work introduces a quantum convolutional encoder-decoder network for pseudo-label generation using Sentinel-1 SAR and Sentinel-2 multispectral imagery, guided only by coarse MODIS-derived land-cover maps. The proposed framework uses quantum-inspired convolutional blocks to enhance feature extraction and cross-sensor fusion, enabling the model to infer a fine-scale semantic structure from heterogeneous inputs. Experiments were conducted on the DFC2020 dataset, which provides a quadruple of Sentinel-1 SAR, Sentinel-2 multispectral, low-resolution MODIS labels and high-resolution reference land-cover maps. The proposed training framework first learns a robust representation of multimodal inputs through the reconstruction branch before any supervision from low-resolution labels is applied. This self-supervised stage enables the model to capture a fine-scale spectral-spatial structure, which is then used to generate high-resolution pseudo-labels. These pseudo-labels are subsequently evaluated through classification metrics on the test data, demonstrating that the proposed approach produces more spatially coherent and class-consistent compared to classical baselines. The results highlight the potential of quantum-inspired deep learning to improve label generation in low-annotation remote sensing scenarios, offering a scalable path towards high-quality land-cover products for EO applications.