Results 21 to 30 of about 25,441 (258)

Dilated convolution with learnable spacings

, 2021
Recent works indicate that convolutional neural networks (CNN) need large receptive fields (RF) to compete with visual transformers and their attention mechanism. In CNNs, RFs can simply be enlarged by increasing the convolution kernel sizes. Yet the number of trainable parameters, which scales quadratically with the kernel's size in the 2D case ...
Khalfaoui-Hassani, Ismail, Pellegrini, Thomas, Masquelier, Timothée   +2 more
openaire   +2 more sources

Smoothed Dilated Convolutions for Improved Dense Prediction [PDF]

Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2018
Dilated convolutions, also known as atrous convolutions, have been widely explored in deep convolutional neural networks (DCNNs) for various dense prediction tasks. However, dilated convolutions suffer from the gridding artifacts, which hampers the performance.
Zhengyang Wang, Shuiwang Ji
openaire   +3 more sources

Inception Convolution with Efficient Dilation Search [PDF]

2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021
As a variant of standard convolution, a dilated convolution can control effective receptive fields and handle large scale variance of objects without introducing additional computational costs. To fully explore the potential of dilated convolution, we proposed a new type of dilated convolution (referred to as inception convolution), where the ...
Liu, Jie, Li, Chuming, Liang, Feng, Lin, Chen, Sun, Ming, Yan, Junjie, Ouyang, Wanli, Xu, Dong   +7 more
openaire   +2 more sources

Application of Three-Dimensional Convolution Network in Brain Hippocampus Segmentation

Jisuanji kexue yu tansuo, 2020
In order to improve the accuracy and robustness of hippocampus segmentation, a new three-dimensional convolutional network named Dilated-3DUnet is proposed.
LIU Chen, XIAO Zhiyong, WU Xinxin
doaj   +1 more source

Dilated Skip Convolution for Facial Landmark Detection [PDF]

Sensors, 2019
Facial landmark detection has gained enormous interest for face-related applications due to its success in facial analysis tasks such as facial recognition, cartoon generation, face tracking and facial expression analysis. Many studies have been proposed and implemented to deal with the challenging problems of localizing facial landmarks from given ...
Seyha Chim, Jin-Gu Lee, Ho-Hyun Park
openaire   +2 more sources

Lightweight image classifier using dilated and depthwise separable convolutions

Journal of Cloud Computing: Advances, Systems and Applications, 2020
The image classification based on cloud computing suffers from difficult deployment as the network depth and data volume increase. Due to the depth of the model and the convolution process of each layer will produce a great amount of calculation, the GPU
Wei Sun, Xiaorui Zhang, Xiaozheng He
doaj   +1 more source

A multiscale dilated convolution and mixed-order attention-based deep neural network for monocular depth prediction

SN Applied Sciences, 2022
Article Highlights We designed an efficient monocular depth prediction framework on the basis of multiscale dilated convolution and a mixed-order attention mechanism. This framework can produce effective depth outputs with rich details.
Huihui Xu, Fei Li
doaj   +1 more source

Convolutions of Harmonic Functions with Certain Dilatations [PDF]

International Journal of Mathematics and Mathematical Sciences, 2017
The convolution of harmonic functions, unlike the analytic case, proved to be very challenging. In this paper, we introduce dilatation conditions that guarantee the convolution of two harmonic functions to be locally one-to-one, sense-preserving, and close-to-convex harmonic in the unit disk.
Om P. Ahuja, Jay M. Jahangiri
openaire   +2 more sources

Dilated Convolution with Dilated GRU for Music Source Separation [PDF]

Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, 2019
Stacked dilated convolutions used in Wavenet have been shown effective for generating high-quality audios. By replacing pooling/striding with dilation in convolution layers, they can preserve high-resolution information and still reach distant locations.
Liu, Jen-Yu, Yang, Yi-Hsuan
openaire   +2 more sources

Remaining Useful Life Prediction of Rolling Bearings Based on Multiscale Convolutional Neural Network with Integrated Dilated Convolution Blocks

Shock and Vibration, 2021
Remaining useful life (RUL) prediction is necessary for guaranteeing machinery’s safe operation. Among deep learning architectures, convolutional neural network (CNN) has shown achievements in RUL prediction because of its strong ability in ...
Ran Wang, Ruyu Shi, Xiong Hu, Changqing Shen   +3 more
doaj   +1 more source