By combining signals from our Imagery, Population and Environment models and datasets, we achieve higher predictive accuracy on real-world classification and forecasting tasks versus a single modality analysis.
多模态模型组合比单一模态分析具有更高的预测准确率，这一发现强调了跨模态融合在提升预测能力方面的重要价值。