Furniture Image Classification Based on Depthwise Group Over-Parameterized Convolution

In this paper, an improved VGG16 combined with depthwise group over-parameterized convolution (DGOVGG16) model is proposed to realize automatic furniture image classification. Firstly, construct convolution, which introduced the VGG 16 for reducing number of parameters overall while extracting more sufficient semantic features images. Then, paper uses ReLU activation function in former part neural network reduce correlation between and accelerate weight update speed model. Meantime, applies Leaky-ReLU last layer avoid problem that some neurons do not update. Compared six classification methods based on MobileNetV2, AlexNet, ShuffleNetv2, GoogleNet, GVGG16, experimental results show DGOVGG16 average accuracy (AA) 95.51% has better performance.