Improved lightweight identification of agricultural diseases based on MobileNetV3

TL;DR

This paper enhances MobileNetV3 with Coordinate Attention to create a lightweight, accurate model for agricultural pest identification, suitable for embedded devices like Jetson Nano.

Contribution

It introduces Coordinate Attention into MobileNetV3, reducing parameters and size while improving accuracy, and demonstrates effective deployment on embedded hardware.

Findings

Parameters reduced by up to 23.4%

Model size decreased by 19.7%

Accuracy increased by up to 2.48% on Jetson Nano

Abstract

At present, the identification of agricultural pests and diseases has the problem that the model is not lightweight enough and difficult to apply. Based on MobileNetV3, this paper introduces the Coordinate Attention block. The parameters of MobileNetV3-large are reduced by 22%, the model size is reduced by 19.7%, and the accuracy is improved by 0.92%. The parameters of MobileNetV3-small are reduced by 23.4%, the model size is reduced by 18.3%, and the accuracy is increased by 0.40%. In addition, the improved MobileNetV3-small was migrated to Jetson Nano for testing. The accuracy increased by 2.48% to 98.31%, and the inference speed increased by 7.5%. It provides a reference for deploying the agricultural pest identification model to embedded devices.