Xu Lin

TL;DR: In this paper, a crack detection method based on depth learning is proposed, which comprises the following steps: step S1, collecting a training set, a verification set and a test set; proportionally andrandomly divided into training set and test set, 2, detecting target classification; the convolution neural network model is trained by transfer learning method.

TL;DR: In this paper, a construction method of binary convolutional neural network suitable for an embedded platform was proposed, which comprises the following steps: S1, collecting an ImageNet data set, and dividing the obtained ImageNet dataset into a training set, a verification set and a test set; S2, according to the obtained training set and verification set, training and verifying the XNOR-Net binarization neural network model to obtain a trained binary network model; S3, integrating the scaling operation and the batch normalization operation in the trained binary networksmodel, and transplant

TL;DR: The BitFlow-Net can remove all floating-point operations in middle layers of BCNNs and greatly reduce the memory for both cases without affecting the accuracy, and can achieve the accuracy comparable to that of the full-precision AlexNet network in the binary classification task.

TL;DR: The receiver operating characteristic curves of BNSC-Net surpass that of the state-of-the-art algorithm DeepIns and are effective and efficient for building deep learning-enabled industrial applications on fog nodes.