A Modified Channel Attention Mechanism For Detecting Diseases In Tea Leaves

Abstract

Detection of diseases in agricultural crop is crucial for improving the quality and yield of agricultural produce. Manual disease detection and identification requires expert guidance and is a costly process. Computing technologies can be highly effective in this endeavour. Convolutional Neural Networks (CNN) are computer vision algorithms that can be used for this purpose. Integration of attention modules with convolutional neural networks can improve the performance of the network. In our study we have integrated channel attention module with MobileNetV2 for detecting diseases in tea leaves. Furthermore, we introduced several modifications to the channel attention mechanism and evaluated their effectiveness in enhancing disease detection performance.