Rotation Invariant Local Binary Convolution Neural Networks

Zhang, Xin; Liu, Li; Xie, Ying; Chen, Jie; Wu, Li-Chung; Pietikäinen, Matti

doi:10.1109/iccvw.2017.146

Cited by 22 publications

(5 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to improve edge performance, researchers have integrated gradient operators into vanilla convolutions, as the original convolution operation tends to smooth local features, resulting in decreased edge sharpness. By utilizing the fixed binary values, which are treated as filters, in convolution instead of learnable kernel weights, local binary convolution (LBC) [ 53 , 54 ] has been explored as an efficient alternative to traditional convolutions in various computer vision tasks. In the context of central difference convolution (CDC) [ 55 , 56 , 57 ], learnable kernels are employed to capture edge and texture details from the central difference map effectively; that is,

, where w indicates the kernel weights, and

represents the surrounding pixel of the center entry in the local patch.…”

Section: Related Workmentioning

confidence: 99%

Semantic Guidance Fusion Network for Cross-Modal Semantic Segmentation

Zhang,

Chen,

Gao

2024

Sensors

View full text Add to dashboard Cite

Leveraging data from various modalities to enhance multimodal segmentation tasks is a well-regarded approach. Recently, efforts have been made to incorporate an array of modalities, including depth and thermal imaging. Nevertheless, the effective amalgamation of cross-modal interactions remains a challenge, given the unique traits each modality presents. In our current research, we introduce the semantic guidance fusion network (SGFN), which is an innovative cross-modal fusion network adept at integrating a diverse set of modalities. Particularly, the SGFN features a semantic guidance module (SGM) engineered to boost bi-modal feature extraction. It encompasses a learnable semantic guidance convolution (SGC) designed to merge intensity and gradient data from disparate modalities. Comprehensive experiments carried out on the NYU Depth V2, SUN-RGBD, Cityscapes, MFNet, and ZJU datasets underscore both the superior performance and generalization ability of the SGFN compared to the current leading models. Moreover, when tested on the DELIVER dataset, the efficiency of our bi-modal SGFN displayed a mIoU that is comparable to the hitherto leading model, CMNEXT.

show abstract

, where w indicates the kernel weights, and

represents the surrounding pixel of the center entry in the local patch.…”

Section: Related Workmentioning

confidence: 99%

Semantic Guidance Fusion Network for Cross-Modal Semantic Segmentation

Zhang,

Chen,

Gao

2024

Sensors

View full text Add to dashboard Cite

show abstract

“…A pioneering work combing LBP and convolutional operations is local binary convolution (LBC) [16,54]. The authors decomposed the pattern generation process of LBP into sub-steps.…”

Section: Local Binary Convolutionmentioning

confidence: 99%

From Local Binary Patterns to Pixel Difference Networks for Efficient Visual Representation Learning

Su¹,

Pietikäinen²,

Liu³

2023

Preprint

View full text Add to dashboard Cite

LBP is a successful hand-crafted feature descriptor in computer vision. However, in the deep learning era, deep neural networks, especially convolutional neural networks (CNNs) can automatically learn powerful task-aware features that are more discriminative and of higher representational capacity. To some extent, such hand-crafted features can be safely ignored when designing deep computer vision models. Nevertheless, due to LBP's preferable properties in visual representation learning, an interesting topic has arisen to explore the value of LBP in enhancing modern deep models in terms of efficiency, memory consumption, and predictive performance. In this paper, we provide a comprehensive review on such efforts which aims to incorporate the LBP mechanism into the design of CNN modules to make deep models stronger. In retrospect of what has been achieved so far, the paper discusses open challenges and directions for future research.

show abstract

“…Section 4 recalls the principles and equations of steerable filters, and explains the roto-translational properties of our filter ensemble. Section 5 describes the proposed architecture, and [17] MNIST-rot n.c. 24 yes no no yes Harmonic networks [18] MNIST-rot 33k Continuous yes no no yes Spherical CNN [19] MNIST/MNIST-rot 68k Continuous yes no yes yes SFCNNs [20] MNIST-rot n.c. 24 yes no yes yes GCNs [21] MNIST-rot 1.86M 4 yes no no yes CIFAR10 Icosahedral CNN [22] MNIST/MNIST-rot n.c. n.c. yes no yes yes RotEqNet [9] MNIST/MNIST-rot 100k 17 yes no yes yes RI-LBCNNs [23] MNIST-rot 390k 8 yes no no yes ORN-8 [24] MNIST/MNIST-rot 969k 8 yes no yes yes CIFAR10 Rot.-Inv. Conv.…”

Section: Paper Organizationmentioning

confidence: 99%

“…The main drawback is high number of parameters to learn. Next, we can cite Rotation Invariant Local Binary Networks [23]. The authors introduce the Local Binary Orientation Module.…”

Section: Internal Architecture Modificationmentioning

confidence: 99%

A minimal model for classification of rotated objects with prediction of the angle of rotation

Salas¹,

Dokládal²,

Dokladalova³

2021

Journal of Visual Communication and Image Representation

View full text Add to dashboard Cite

In classification tasks, the robustness against various image transformations remains a crucial property of the Convolutional Neural Networks (CNNs). It can be acquired using the data augmentation. It comes, however, at the price of the risk of overfitting and a considerable increase in training time. Consequently, other ways to endow CNN with invariance to various transformations-and mainly to the rotations-is an intensive field of study. This paper presents a new reduced rotation invariant classification model composed of two parts: a feature representation mapping and a classifier. We provide an insight into the principle and we prove that the proposed model is trainable. This model is smaller in terms of trainable parameters than similar approaches, and has angular prediction capabilities. We illustrate the results on the MNIST and CIFAR-10 datasets. On MNIST, we i) achieve the state of the art of classification on MNIST-rot (with training on MNIST-rot), and ii) improve the results of classification on MNIST-rot (with training on upright MNIST). When trained on CIFAR-10 with upright samples and tested with rotated samples we improve by 20% the state of the art classification results. In all cases, we can predict the rotation angle.

show abstract

Rotation Invariant Local Binary Convolution Neural Networks

Cited by 22 publications

References 21 publications

Semantic Guidance Fusion Network for Cross-Modal Semantic Segmentation

Semantic Guidance Fusion Network for Cross-Modal Semantic Segmentation

From Local Binary Patterns to Pixel Difference Networks for Efficient Visual Representation Learning

A minimal model for classification of rotated objects with prediction of the angle of rotation

Contact Info

Product

Resources

About