paper-with-me

홈 › Papers

Orthogonal Convolutional Neural Networks

2019-11-27 · CVPR 2020 6 · Jiayun Wang, Yubei Chen, Rudrasis Chakraborty, Stella X. Yu

Deep convolutional neural networks are hindered by training instability and feature redundancy towards further performance improvement. A promising solution is to impose orthogonality on convolutional filters. We develop an efficient approach to impose filter orthogonality on a convolutional layer based on the doubly block-Toeplitz matrix representation of the convolutional kernel instead of using the common kernel orthogonality approach, which we show is only necessary but not sufficient for ensuring orthogonal convolutions. Our proposed orthogonal convolution requires no additional parameters and little computational overhead. This method consistently outperforms the kernel orthogonality alternative on a wide range of tasks such as image classification and inpainting under supervised, semi-supervised and unsupervised settings. Further, it learns more diverse and expressive features with better training stability, robustness, and generalization. Our code is publicly available at https://github.com/samaonline/Orthogonal-Convolutional-Neural-Networks.

📄 PDF Abstract BibTeX arXiv:1911.12207

Code (1)

samaonline/Orthogonal-Convolutional-Neural-Networks 공식 구현 pytorch

Tasks

image-classificationImage ClassificationImage Retrieval

Methods 이 논문이 사용한 방법론

Convolution A convolution is a type of matrix operation, consisting of a kernel, a small matrix of weights, that slides over input data performing element-wise multiplication with the…

Similar Papers 제목 키워드 기반

Scaling-up Diverse Orthogonal Convolutional Networks with a Paraunitary Framework

2021-06-16 · Jiahao Su, Wonmin Byeon, Furong Huang

Enforcing orthogonality in neural networks is an antidote for gradient vanishing/exploding problems, sensitivity by adversarial perturbation, and bounding generalization errors. However, many previous approaches are heur…

Scaling-up Diverse Orthogonal Convolutional Networks by a Paraunitary Framework

2021-09-29 · Jiahao Su, Wonmin Byeon, Furong Huang

Enforcing orthogonality in neural networks is an antidote for gradient vanishing/exploding problems, sensitivity to adversarial perturbation, and bounding generalization errors. However, many previous approaches are heur…

Existence, Stability and Scalability of Orthogonal Convolutional Neural Networks

2021-08-12 · El Mehdi Achour, François Malgouyres, Franck Mamalet

Imposing orthogonality on the layers of neural networks is known to facilitate the learning by limiting the exploding/vanishing of the gradient; decorrelate the features; improve the robustness. This paper studies the th…

Orthogonal Decomposition Network for Pixel-Wise Binary Classification

2019-06-01 · CVPR 2019 6 · Chang Liu, Fang Wan, Wei Ke, Zhuowei Xiao 외

The weight sharing scheme and spatial pooling operations in Convolutional Neural Networks (CNNs) introduce semantic correlation to neighboring pixels on feature maps and therefore deteriorate their pixel-wise classificat…

Binary ClassificationClassificationEdge DetectionGeneral Classification+2

Orthogonalizing Convolutional Layers with the Cayley Transform

2021-04-14 · ICLR 2021 1 · Asher Trockman, J. Zico Kolter

Recent work has highlighted several advantages of enforcing orthogonality in the weight layers of deep networks, such as maintaining the stability of activations, preserving gradient norms, and enhancing adversarial robu…

Adversarial Robustness