An Improved Deep Fusion CNN for Image Recognition

With the development of Deep Convolutional Neural Networks(DCNNs),the extracted features for image recognition tasks have shifted from low-level features to the high-level semantic features of DCNNs.Previous studies have shown that the deeper the network is,the more abstract the features are.However,the recognition ability of deep features would be limited by insufficient training samples.To address this problem,this paper derives an improved Deep Fusion Convolutional Neural Network(DF-Net)which can make full use of the differences and complementarities during network learning and enhance feature expression under the condition of limited datasets.Specifically,DF-Net organizes two identical subnets to extract features from the input image in parallel,and then a well-designed fusion module is introduced to the deep layer of DF-Net to fuse the subnet’s features in multi-scale.Thus,the more complex mappings are created and the more abundant and accurate fusion features can be extracted to improve recognition accuracy.Furthermore,a corresponding training strategy is also proposed to speed up the convergence and reduce the computation overhead of network training.Finally,DF-Nets based on the well-known ResNet,DenseNet and MobileNetV2 are evaluated on CIFAR100,Stanford Dogs,and UECFOOD-100.Theoretical analysis and experimental results strongly demonstrate that DF-Net enhances the performance of DCNNs and increases the accuracy of image recognition.

An Effective Image-Based Tomato Leaf Disease Segmentation Method Using MC-UNet

Tomato disease control is an urgent requirement in the field of intellectual agriculture,and one of the keys to it is quantitative identification and precise segmentation of tomato leaf diseases.Some diseased areas on tomato leaves are tiny and may go unnoticed during segmentation.Blurred edge also makes the segmentation accuracy poor.Based on UNet,we propose an effective image-based tomato leaf disease segmentation method called Cross-layer Attention Fusion Mechanism combined with Multi-scale Convolution Module(MC-UNet).First,a Multi-scale Convolution Module is proposed.This module obtains multiscale information about tomato disease by employing 3 convolution kernels of different sizes,and it highlights the edge feature information of tomato disease using the Squeeze-and-Excitation Module.Second,a Cross-layer Attention Fusion Mechanism is proposed.This mechanism highlights tomato leaf disease locations via gating structure and fusion operation.Then,we employ SoftPool rather than MaxPool to retain valid information on tomato leaves.Finally,we use the SeLU function appropriately to avoid network neuron dropout.We compared MC-UNet to the existing segmentation network on our self-built tomato leaf disease segmentation dataset and MC-UNet achieved 91.32%accuracy and 6.67M parameters.Our method achieves good results for tomato leaf disease segmentation,which demonstrates the effectiveness of the proposed methods.