Convolutional Neural Networks with Alternately Updated Clique

Yang, Yibo; Zhong, Zhisheng; Shen, Tiantian; Lin, Zhouchen

doi:10.1109/cvpr.2018.00256

Cited by 134 publications

(111 citation statements)

References 29 publications

Supporting

Mentioning

111

Contrasting

Order By: Relevance

“…Semantic segmentation. Fully convolutional network (FCN) [22] based methods have made great progress in image semantic segmentation by leveraging the powerful convolutional features of classification networks [14,15,33] pre-trained on large-scale data [28]. Several model variants are proposed to enhance the multi-scale contextual aggregation.…”

Section: Related Workmentioning

confidence: 99%

Expectation-Maximization Attention Networks for Semantic Segmentation

Li¹,

Zhong

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

Self Cite

568

300

View full text Add to dashboard Cite

Self-attention mechanism has been widely used for various tasks. It is designed to compute the representation of each position by a weighted sum of the features at all positions. Thus, it can capture long-range relations for computer vision tasks. However, it is computationally consuming. Since the attention maps are computed w.r.t all other positions. In this paper, we formulate the attention mechanism into an expectation-maximization manner and iteratively estimate a much more compact set of bases upon which the attention maps are computed. By a weighted summation upon these bases, the resulting representation is low-rank and deprecates noisy information from the input. The proposed Expectation-Maximization Attention (EMA) module is robust to the variance of input and is also friendly in memory and computation. Moreover, we set up the bases maintenance and normalization methods to stabilize its training procedure. We conduct extensive experiments on popular semantic segmentation benchmarks including PAS-CAL VOC, PASCAL Context and COCO Stuff, on which we set new records 1 .

show abstract

Section: Related Workmentioning

confidence: 99%

Expectation-Maximization Attention Networks for Semantic Segmentation

Li¹,

Zhong

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

Self Cite

568

300

View full text Add to dashboard Cite

show abstract

“…The initial learning rate is usually the single most important hyper-parameter, we should always make sure that it is tuned [32]. There are two structure parameters to set first [26], T is the sum of the layers of all the blocks, k represents the number of filters per layer in each block. The kernel size of convolution layers in all blocks are 3 × 3, and we use one-pixel padding to keep the size of the matrix the same before and after the convolution.…”

Section: Training Detailsmentioning

confidence: 99%

Automatic Tunnel Crack Detection Based on U-Net and a Convolutional Neural Network with Alternately Updated Clique

et al. 2020

Sensors

View full text Add to dashboard Cite

Regular crack inspection of tunnels is essential to guarantee their safe operation. At present, the manual detection method is time-consuming, subjective and even dangerous, while the automatic detection method is relatively inaccurate. Detecting tunnel cracks is a challenging task since cracks are tiny, and there are many noise patterns in the tunnel images. This study proposes a deep learning algorithm based on U-Net and a convolutional neural network with alternately updated clique (CliqueNet), called U-CliqueNet, to separate cracks from background in the tunnel images. A consumer-grade DSC-WX700 camera (SONY, Wuxi, China) was used to collect 200 original images, then cracks are manually marked and divided into sub-images with a resolution of 496 × 496 pixels. A total of 60,000 sub-images were obtained in the dataset of tunnel cracks, among which 50,000 were used for training and 10,000 were used for testing. The proposed framework conducted training and testing on this dataset, the mean pixel accuracy (MPA), mean intersection over union (MIoU), precision and F1-score are 92.25%, 86.96%, 86.32% and 83.40%, respectively. We compared the U-CliqueNet with fully convolutional networks (FCN), U-net, Encoder–decoder network (SegNet) and the multi-scale fusion crack detection (MFCD) algorithm using hypothesis testing, and it’s proved that the MIoU predicted by U-CliqueNet was significantly higher than that of the other four algorithms. The area, length and mean width of cracks can be calculated, and the relative error between the detected mean crack width and the actual mean crack width ranges from −11.20% to 18.57%. The results show that this framework can be used for fast and accurate crack semantic segmentation of tunnel images.

show abstract

“…Batch normalization [31] can accelerate deep network training by reducing internal covariate shift. It allows us to use much higher learning rates and be less careful about initialization.…”

Section: Octonion Batch Normalization Modulementioning

confidence: 99%

Deep octonion networks

et al. 2020

Neurocomputing

View full text Add to dashboard Cite

Deep learning is a research hot topic in the field of machine learning. Real-value neural networks (Real NNs), especially deep real networks (DRNs), have been widely used in many research fields. In recent years, the deep complex networks (DCNs) and the deep quaternion networks (DQNs) have attracted more and more attentions. The octonion algebra, which is an extension of complex algebra and quaternion algebra, can provide more efficient and compact expression. This paper constructs a general framework of deep octonion networks (DONs) and provides the main building blocks of DONs such as octonion convolution, octonion batch normalization and octonion weight initialization; DONs are then used in image classification tasks for CIFAR-10 and CIFAR-100 data sets. Compared with the DRNs, the DCNs, and the DQNs, the proposed DONs have better convergence and higher classification accuracy. The success of DONs is also explained by multi-task learning. IntroductionReal-value neural networks (Real NNs) [1][2][3][4][5][6][7][8][9][10][11][12] attracted the attention of many researchers and recently made major breakthroughs in many areas such as signal processing, image processing, natural language processing, etc.Many models of Real NNs have been constructed in the literature. These models can generally be categorized into two kinds: non-deep models and deep models. The non-deep models are mainly constructed by multilayer perceptron module [13] and hard to train, if we only use the real-valued back propagation (BP) algorithm [14], when their layers are larger than 4. The deep models can be roughly constructed by the following two strategies: multilayer perceptron models assisted by the unsupervised pretrained methods (for example, deep belief nets [15], deep auto-encoder [16], etc.) and real-value convolutional neural networks (Real CNNs), including LeNet-5 [17], AlexNet [18], Inception [19-22], VGGNet [23], HighwayNet [24], ResNet [25], ResNeXt [26], DenseNet [27], FractalNet [28], PolyNet [29], SENet [30], CliqueNet [31], BinaryNet [32], SqueezeNet [33], MobileNet [34], etc.Although Real CNNs have achieved great success in various applications, the correlations between convolution kernels generally do not take into consideration, that is, there are no connections or no special relationships considering between convolution kernels. The opposite of Real CNNs is real-value recurrent neural networks (Real RNNs) [35][36][37][38], who obtain the correlations by adding the connections between convolution kernels and then learn the weights of these connections, which, however, increased significantly the training difficulty and was easier to encounter converge problems. The first question has been raised: Can we consider the correlations between convolution kernels by some special relationships, which do not need to learn, instead of adding the connections between convolution kernels?Many researchers find that the performance can be improved when the relationships between convolution kernels are modeled by complex algebra, quaterni...

show abstract

Convolutional Neural Networks with Alternately Updated Clique

Cited by 134 publications

References 29 publications

Expectation-Maximization Attention Networks for Semantic Segmentation

Expectation-Maximization Attention Networks for Semantic Segmentation

Automatic Tunnel Crack Detection Based on U-Net and a Convolutional Neural Network with Alternately Updated Clique

Deep octonion networks

Contact Info

Product

Resources

About