Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces CascadeCropNet, a two-stage cascade architecture designed for crop damage assessment in agricultural insurance, specifically for smallholder farmers. By implementing a lightweight Sentinel model for initial health triage followed by an Expert model for detailed diagnosis, the system achieves a target recall of 0.974 while effectively managing the trade-off between missed detections and false alarms. Evaluated on a dataset of 23,804 images, the cascade architecture reduces missed damage cases by up to 54% compared to traditional classifiers, demonstrating its potential for enhancing operational safety in agricultural assessments.
Achieving a 0.974 recall in crop damage detection while cutting missed cases by 54% could revolutionize agricultural insurance for smallholder farmers.
In picture-based agricultural insurance for smallholder farmers, missed damage detections carry substantially higher cost than false alarms: a farmer who sustained real losses receives no payout, while unnecessary expert review is operationally costly but reversible. Standard multi-class classifiers optimize global accuracy but provide no mechanism to operationalize or control this asymmetric cost structure at inference time. We propose CascadeCropNet, a two-stage cascade architecture calibrated to satisfy a target recall constraint (Rec-Damaged>= 0.95) through threshold selection. A lightweight Sentinel model performs binary health triage; samples exceeding a calibrated damage probability threshold tau are escalated to a specialist Expert model for fine-grained diagnosis. This design provides explicit, deployment-time control over the safety-efficiency trade-off without retraining. Evaluated on the Eyes on the Ground dataset (23,804 images from Kenyan smallholder maize farms), the cascade achieves Rec-Damaged = 0.974 at tau = 0.5, reducing missed damage cases by up to 54% relative to a flat baseline. Under evaluation alignment, the representational gap reduces to +0.008 F1-macro, confirming the contribution is architectural rather than representational. Under input degradation, the system prioritizes escalation over confident misclassification, reflecting error containment through architectural isolation rather than intrinsic model robustness. These results demonstrate that cascade architectures can operationalize safety-oriented decision constraints through calibrated routing in settings where reliability matters more than aggregate accuracy. These properties depend on threshold calibration and deployment conditions and do not constitute guarantees under arbitrary distribution shift.