ass日本风韵熟妇pics男人扒开女人屁屁桶到爽|扒开胸露出奶头亲吻视频|邻居少妇的诱惑|人人妻在线播放|日日摸夜夜摸狠狠摸婷婷|制服 丝袜 人妻|激情熟妇中文字幕|看黄色欧美特一级|日本av人妻系列|高潮对白av,丰满岳妇乱熟妇之荡,日本丰满熟妇乱又伦,日韩欧美一区二区三区在线

In order to explore the effect of depth of convolution layer on feature extraction of kiwi trunk images, a visualization method was proposed to analyze the extracted features. Firstly, the collected data set was classified into positive and negative samples. Taking the area where the trunk and the water pipe intersected in the dataset as positive samples and the remaining areas as negative samples. Input the samples into LeNet, Alexnet, Vgg-16 and the defined three types of shallow structures for training. Then, by extracting the activation map, normalization, and bicubic interpolation visualization methods, the visualization results of the last convolution layer of each classification model were obtained. The comparison can be obtained: Alexnet and Vgg-16 extracted trunk features in the test image, while LeNet and three types of shallow models extracted the trunk, ridge and other features together while extracting the trunk. Finally, the image classification and object detection models of the above six types of network structures as feature extraction layers were used to verify the flowering period and fruiting period data sets, and the accuracy drop caused by changes in the characteristics of the data sets in different seasons was used as the evaluation criterion: the accuracy of image classification shallow model was decreased by more than 15.90 percentage points, Alexnet and Vgg-16 were decreased by 6.94 percentage points and 2.08 percentage points respectively, the accuracy of object detection shallow model was decreased by more than 49.77 percentage points, Alexnet and Vgg-16 were decreased by 22.53 percentage points and 20.54 percentage points respectively. The accuracy of all shallow models was greatly reduced due to changes in the extracted features. This method explained the difference between the feature extraction of the kiwi trunk target from the deep network and the shallow network from the perspective of visualization, and provided a reference for the adjustment of network depth and training samples in subsequent research. 