A Novel Optimized Deep Learning Model for Canola Crop Yield Prediction on Edge Devices

Abstract

The escalating global demand for food, coupled with challenges in sustaining crop production, deteriorating ocean health, and depleting natural resources, underscores the critical role of agricultural technology. This article addresses the imperative of developing an optimal deep-learning model for predicting canola crop yield using hyperspectral images captured by drone flights. Our primary objective is to identify the most efficient model in terms of performance and size, considering the storage limitations on edge devices like Raspberry Pi 4 (RPi4). We start with the baseline 1D_CNN model, which achieves an$R^{2}$score of 0.82, and compress it into the proposedfs_model(fp32). To achieve the compression, we apply pruning through sparsity and feature selection using SHAP values. Further reduction in model size is accomplished by quantizing the weights of the proposed model to a lower precision, such as int16. This combined approach substantially decreases the proposed model's size by approximately 92.6% and inference time by approximately ×9013 in comparison to the baseline 1D_CNN model. In addition, we propose the novelfsp_modelposit(8,3) that uses posit quantization to further reduce the computation requirements compared to the proposedfs_model(int16). Our findings indicate that the utilization of posit numbers enables us to shrink the model size to 94% of the original base model, while only reducing the$R^{2}$score by 5.7%.

Plant phenotyping relevance

ドローンのハイパースペクトル画像からカノーラ収量を推定する深層学習モデルの圧縮・量子化・性能評価が研究の中心であり、植物形質の取得・推定手法に該当する。

Code and data availability

公開本文の所在を確認できませんでした。非公開または購読が必要な可能性があります。

No evidence-backed public reproduction asset is currently recorded.