Enhancing inference speed in reparameterized convolutional neural network for vibration-based damage detection

Di Wang, Yuanming Lu, Xiangli Yang, Die Liu, Xianyi Yang, Jianxi Yang

Abstract: Highlights•A convolutional neural network based on the reparameterization technique has been introduced, which can improve the efficiency of VBDD tasks.•This network is a VGG-style network with no branches.•Introducing a new zero-padding method, which enables rectangular convolution kernels to be applied to the reparameterization technique.