Convolutional Kernel

Convolutional kernels are the fundamental building blocks of convolutional neural networks (CNNs), responsible for extracting features from data by applying weighted sums across local regions. Current research focuses on improving kernel design for efficiency and effectiveness, exploring techniques like dynamic kernels (adapting to input data), multi-kernel approaches (capturing diverse features), and specialized kernels for specific data types (e.g., time series, 3D point clouds). These advancements are driving improvements in various applications, including image processing, speech recognition, and time series analysis, by enhancing model accuracy, reducing computational costs, and improving interpretability.

Papers

June 6, 2022

Stacked unsupervised learning with a network architecture found by supervised meta-learning
Kyle Luther, H. Sebastian Seung
Unsupervised Learning Back Propagation Convolutional Kernel Unsupervised Algorithm Network Architecture Nondeterministic Stack

May 13, 2022

Blind Image Inpainting with Sparse Directional Filter Dictionaries for Lightweight CNNs
Jenny Schmalfuss, Erik Scheurer, Heng Zhao, Nikolaos Karantzas, Andrés Bruhn, Demetrio Labate
Convolutional Neural Network Deep Learning Architecture Convolutional Kernel Lightweight CNN Blind Image

April 12, 2022

Speech Emotion Recognition with Global-Aware Fusion on Multi-scale Feature Representation
Wenjing Zhu, Xiang Li
Speech Emotion Recognition Multi Scale Feature Convolutional Kernel Multi Scale Feature Representation Context Aware Cross Level Fusion

March 25, 2022

HYDRA: Competing convolutional kernels for fast and accurate time series classification
Angus Dempster, Daniel F. Schmidt, Geoffrey I. Webb
Time Series Time Series Classification Convolutional Kernel Hydra MDP Input Time Series Dictionary Based

March 10, 2022

Compressing CNN Kernels for Videos Using Tucker Decompositions: Towards Lightweight CNN Applications
Tobias Engelhardt Rasmussen, Line H Clemmensen, Andreas Baum
Neural Network Convolutional Neural Network Gameplay Video CNN Model Direct Convolution Convolutional Filter Convolutional Kernel Tucker Decomposition

March 7, 2022

S-Rocket: Selective Random Convolution Kernels for Time Series Classification
Hojjat Salehinejad, Yang Wang, Yuanhao Yu, Tang Jin, Shahrokh Valaee
Time Series Classification Ridge Regression Convolutional Kernel Convolution Operation Random Convolution Kernel Trick

March 3, 2022

Symmetry Structured Convolutional Neural Networks
Kehelwala Dewage Gayan Maduranga, Vasily Zadorozhnyy, Qiang Ye
Convolutional Neural Network Convolutional Layer CNN Architecture Convolutional Kernel Symmetric Network

February 15, 2022

DualConv: Dual Convolutional Kernels for Lightweight Deep Neural Networks
Jiachen Zhong, Junying Chen, Ajmal Mian
Neural Network CNN Architecture Convolutional Kernel Lightweight Deep Dual Convolutional

February 6, 2022

Hyper-Convolutions via Implicit Kernels for Medical Imaging
Tianyu Ma, Alan Q. Wang, Adrian V. Dalca, Mert R. Sabuncu
Convolutional Neural Network Medical Imaging Convolutional Kernel Implicit Kernel Hyper Convolution

January 13, 2022

Largest Eigenvalues of the Conjugate Kernel of Single-Layered Neural Networks
Lucas Benigni, Sandrine Péché
Convolutional Kernel Layer Neural Network Largest Eigenvalue

January 2, 2022

Fast and High-Quality Image Denoising via Malleable Convolutions
Yifan Jiang, Bartlomiej Wronski, Ben Mildenhall, Jonathan T. Barron, Zhangyang Wang, Tianfan Xue
Image Denoising Deformable Convolution Convolutional Kernel Dynamic Convolution Spatial Convolution

November 16, 2021

Convolutional Kernel

Papers

Stacked unsupervised learning with a network architecture found by supervised meta-learning

Blind Image Inpainting with Sparse Directional Filter Dictionaries for Lightweight CNNs

Speech Emotion Recognition with Global-Aware Fusion on Multi-scale Feature Representation

HYDRA: Competing convolutional kernels for fast and accurate time series classification

Compressing CNN Kernels for Videos Using Tucker Decompositions: Towards Lightweight CNN Applications

S-Rocket: Selective Random Convolution Kernels for Time Series Classification

Symmetry Structured Convolutional Neural Networks

DualConv: Dual Convolutional Kernels for Lightweight Deep Neural Networks

Hyper-Convolutions via Implicit Kernels for Medical Imaging

Largest Eigenvalues of the Conjugate Kernel of Single-Layered Neural Networks

Fast and High-Quality Image Denoising via Malleable Convolutions

Image-specific Convolutional Kernel Modulation for Single Image Super-resolution

Learning with convolution and pooling operations in kernel methods