Urban village recognition based on multimodal data fusion with an interpretable graph neural network

Key Points

• Recognition accuracy improves with multimodal data integration, enhancing urban village identification.
• Key evidence shows 20% increase in recognition accuracy when using graph neural networks and multimodal data.
• Analysis employs advanced machine learning techniques to fuse diverse data sources for urban recognition tasks.
• This method may enable smarter urban planning but requires validation across different city environments.