RS-CapsNet: An Advanced Capsule Network

Yang, Shuai; Lee, Feifei; Miao, Ran; Cai, Jiawei; Chen, Lu; Wei, Yuan; Kotani, Kazuhiko; Qiu, Chengfeng

doi:10.1109/access.2020.2992655

Cited by 43 publications

(21 citation statements)

References 27 publications

(30 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Later, Hinton et al published Matrix Capsules, a capsule network with a more powerful routing algorithm based on expectation-maximization and reported a new state-of-theart performance on the Small-NORB dataset. Many variants of capsule networks have been proposed in recent years ( [21], [29], [30], [3], [4], [25], [18]), but for many tasks (i.e., CIFAR-10, CIFAR-100 or ImageNet) even the most recent approaches [21] provide much higher error rates the than state-of-the-art CNN approaches (i.e., 9% on CIFAR-10, while CNNs provide errors of just 3% [26]). Additionally, capsule networks require vast amounts of memory.…”

Section: Capsule Networkmentioning

confidence: 99%

Momentum Capsule Networks

Gugglberger¹,

Peer²,

Rodríguez-Sánchez³

2022

Preprint

View full text Add to dashboard Cite

Capsule networks are a class of neural networks that achieved promising results on many computer vision tasks. However, baseline capsule networks have failed to reach state-of-the-art results on more complex datasets due to the high computation and memory requirements. We tackle this problem by proposing a new network architecture, called Momentum Capsule Network (MoCapsNet). MoCapsNets are inspired by Momentum ResNets, a type of network that applies reversible residual building blocks. Reversible networks allow for recalculating activations of the forward pass in the backpropagation algorithm, so those memory requirements can be drastically reduced. In this paper, we provide a framework on how invertible residual building blocks can be applied to capsule networks. We will show that MoCapsNet beats the accuracy of baseline capsule networks on MNIST, SVHN and CIFAR-10 while using considerably less memory. The source code is available on https://github.com/moejoe95/MoCapsNet.

show abstract

Section: Capsule Networkmentioning

confidence: 99%

Momentum Capsule Networks

Gugglberger¹,

Peer²,

Rodríguez-Sánchez³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…The traditional CapsNet is insufficient in the deep feature extraction capability. In the field of remote sensing image application, researchers have improved the feature extractor of the capsule network, such as residual capsule network (Res-CapsNet) [21], densely connected capsule network (Dense-CapsNet) [26], and CNN-CapsNet [28]. We apply these methods to the land cover classification of PolSAR images and compare them with the proposed method.…”

Section: Comparison Of Different Feature Extractorsmentioning

confidence: 99%

“…In recent years, the capsule network (CapsNet) has been applied to many image applications and it has obtained more competitive results than CNNs [20][21][22]. Sabour and Hinton et al [23] proposed the CapsNet, which encapsulates multiple neurons into a vector.…”

Section: Introductionmentioning

confidence: 99%

PolSAR Image Land Cover Classification Based on Hierarchical Capsule Network

et al. 2021

View full text Add to dashboard Cite

Polarimetric synthetic aperture radar (PolSAR) image classification is one of the basic methods of PolSAR image interpretation. Deep learning algorithms, especially convolutional neural networks (CNNs), have been widely used in PolSAR image classification due to their powerful feature learning capabilities. However, a single neuron in the CNN cannot represent multiple polarimetric attributes of the land cover. The capsule network (CapsNet) uses vectors instead of the single neuron to characterize the polarimetric attributes, which improves the classification performance compared with traditional CNNs. In this paper, a hierarchical capsule network (HCapsNet) is proposed for the land cover classification of PolSAR images, which can consider the deep features obtained at different network levels in the classification. Moreover, we adopt three attributes to uniformly describe the scattering mechanisms of different land covers: phase, amplitude, and polarimetric decomposition parameters, which improves the generalization performance of HCapsNet. Furthermore, conditional random field (CRF) is added to the classification framework to eliminate small isolated regions of the intra-class. Comprehensive evaluations are performed on three PolSAR datasets acquired by different sensors, which demonstrate that our proposed method outperforms other state-of-the-art methods.

show abstract

“…In convolutional layer, the capsule is obtained based on the feature maps that formed by the convolution operation of the input images. Many improved CapsNets add new convolution modules, such as the introduction of DenseNet [15], Res2Net [23] and U-Net [24], to enhance the feature extraction. Besides, the design of multiscale method can also increase the effect of the feature extraction, C. Xiang et al [25] use multi-scale feature extraction to transform the feature maps into different dimensions of the primary capsule.…”

Section: A Capsule Networkmentioning

confidence: 99%

An Improved Capsule Network Based on Capsule Filter Routing

et al. 2021

Self Cite

View full text Add to dashboard Cite

Capsule network (CapsNet) is a novel type of network that can retain spatial information, because each capsule can integrate more information than scalar-output features. However, the CapsNet learns all the features in the input image due to the lack of pooling operation, and there is no connection between different layers in the multi-layer network structure. In this paper, we propose an improved capsule network (CapsNet) based on Capsule Filter Routing (CFR) to address this problem, namely CFR-CapsNet. Firstly, we propose a new routing method called CFR for filtering capsules based on capsule activation values, which can speed up the operation of the model, and then introduce a self-attention mechanism to improve the performance of the primary capsule in the capsule space. Furthermore, in the multi-layer network structure, we transmit the information of the classified capsule with the largest activation value in the previous capsule layer to the primary capsules of the next layer, which improves the relevance of the overall structure. We conduct experiments on Fashion-MNIST, SVHN, and CIFAR-10/100, and the experimental results show that our method can improve the performance of the CapsNet more effectively than other state-of-the-arts.

show abstract

RS-CapsNet: An Advanced Capsule Network

Cited by 43 publications

References 27 publications

Momentum Capsule Networks

Momentum Capsule Networks

PolSAR Image Land Cover Classification Based on Hierarchical Capsule Network

An Improved Capsule Network Based on Capsule Filter Routing

Contact Info

Product

Resources

About