Problem
Research questions and friction points this paper is trying to address.
Improves corn yield prediction accuracy
Addresses mixed pixel issue
Utilizes pixel-level data analysis
Learning county from pixels: Corn yield prediction with attention-weighted multiple instance learning
🤖 AI Summary
Pixel-level aggregation in county-scale maize yield prediction discards spatial heterogeneity, while resolution mismatches between remote sensing feature maps and crop masks introduce mixed-pixel noise. Method: We propose an Attention-Weighted Multi-Instance Learning (AW-MIL) framework that models spatial variability directly at the pixel level—bypassing conventional county-level averaging—and integrates attention mechanisms into MIL for adaptive suppression of mixed-pixel interference. Contribution/Results: Validated spatiotemporally, AW-MIL significantly improves prediction robustness and interpretability. Over five years (2018–2022) across the U.S. Corn Belt, it achieves an R² of 0.84 and RMSE of 0.83 in 2022—outperforming four state-of-the-art methods. Attention weight visualizations confirm its capacity to focus on critical growth regions while effectively filtering out noise, demonstrating both quantitative superiority and qualitative interpretability.
📝 Abstract
Remote sensing technology has become a promising tool in yield prediction. Most prior work employs satellite imagery for county-level corn yield prediction by spatially aggregating all pixels within a county into a single value, potentially overlooking the detailed information and valuable insights offered by more granular data. To this end, this research examines each county at the pixel level and applies multiple instance learning to leverage detailed information within a county. In addition, our method addresses the"mixed pixel"issue caused by the inconsistent resolution between feature datasets and crop mask, which may introduce noise into the model and therefore hinder accurate yield prediction. Specifically, the attention mechanism is employed to automatically assign weights to different pixels, which can mitigate the influence of mixed pixels. The experimental results show that the developed model outperforms four other machine learning models over the past five years in the U.S. corn belt and demonstrates its best performance in 2022, achieving a coefficient of determination (R2) value of 0.84 and a root mean square error (RMSE) of 0.83. This paper demonstrates the advantages of our approach from both spatial and temporal perspectives. Furthermore, through an in-depth study of the relationship between mixed pixels and attention, it is verified that our approach can capture critical feature information while filtering out noise from mixed pixels.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Pixel-level county analysis
Attention mechanism weighting
Mixed pixel noise reduction
🔎 Similar Papers
No similar papers found.
