Efficient CNN

Efficient Convolutional Neural Networks (CNNs) aim to minimize computational cost (FLOPs, parameters, latency) while maintaining high accuracy for computer vision tasks. Current research focuses on novel architectures like Columnar Stage Networks and ShuffleMixers, as well as optimization techniques such as filter pruning, block fusion, and efficient modulation mechanisms, to achieve this balance. These advancements are crucial for deploying CNNs on resource-constrained devices (e.g., mobile phones, embedded systems) and for improving the scalability of large-scale applications.

Papers

October 5, 2024

Designing Concise ConvNets with Columnar Stages
Ashish Kumar, Jaesik Park
Convolution Layer High Performing 1D ConvNet Multiple Stage Efficient CNN Stage Network

April 4, 2024

On the Efficiency of Convolutional Neural Networks
Andrew Lavin
Convolutional Neural Network High Efficiency High Performing 1D ConvNet Computational Efficiency Efficient CNN

March 29, 2024

Efficient Modulation for Vision Networks
Xu Ma, Xiyang Dai, Jianwei Yang, Bin Xiao, Yinpeng Chen, Yun Fu, Lu Yuan
Efficient CNN Efficient Network Modulation Scheme Visual Network

January 12, 2024

UPDP: A Unified Progressive Depth Pruner for CNN and Vision Transformer
Ji Liu, Dehua Tang, Yuanxian Huang, Li Zhang, Xiaocheng Zeng, Dong Li, Mingjie Lu, Jinzhang Peng, Yu Wang, Fan Jiang, Lu Tian, Ashish Sirasao
Vision Transformer CNN Network Channel Pruning Depth Wise Convolution Efficient CNN Depth Pruning

July 1, 2023

Filter Pruning for Efficient CNNs via Knowledge-driven Differential Filter Sampler
Shaohui Lin, Wenxuan Huang, Jiao Xie, Baochang Zhang, Yunhang Shen, Zhou Yu, Jungong Han, David Doermann
Feature Redundancy Filter Pruning Efficient CNN Source Filter Mask Propagation

February 4, 2023

Revisiting Image Deblurring with an Efficient ConvNet
Lingyan Ruan, Mojtaba Bemana, Hans-peter Seidel, Karol Myszkowski, Bin Chen
Convolutional Neural Network Receptive Field Video Deblurring Convolutional Block Depth Wise Convolution Efficient CNN

November 9, 2022

Efficient Large-scale Audio Tagging via Transformer-to-CNN Knowledge Distillation
Florian Schmid, Khaled Koutini, Gerhard Widmer
Knowledge Distillation Audio Spectrogram Transformer Efficient CNN Audio Tagging Complex Transformer

September 22, 2022

Efficient CNN with uncorrelated Bag of Features pooling
Firas Laakom, Jenni Raitoharju, Alexandros Iosifidis, Moncef Gabbouj
Feature Wise Convolution Layer Bag of Word Histogram Layer Efficient CNN Pairwise Correlation

August 14, 2022

Remote Photoplethysmography from Low Resolution videos: An end-to-end solution using Efficient ConvNets
Bharath Ramakrishnan, Ruijia Deng
Deep Learning Convolutional Neural Network Deep Neural Network End to End Remote Photoplethysmography Facial Video Efficient CNN Low Resolution Video

July 29, 2022

Low-complexity Approximate Convolutional Neural Networks
R. J. Cintra, S. Duffner, C. Garcia, A. Leite
Convolutional Neural Network High Performing 1D ConvNet Convolutional Filter Low Complexity Efficient CNN

May 30, 2022

ShuffleMixer: An Efficient ConvNet for Image Super-Resolution
Long Sun, Jinshan Pan, Jinhui Tang
Super Resolution Image Super Resolution Convolution Kernel Efficient CNN

May 23, 2022

Training Efficient CNNS: Tweaking the Nuts and Bolts of Neural Networks for Lighter, Faster and Robust Models
Sabeesh Ethiraj, Bharath Kumar Bolla
Neural Network Deep Learning Convolutional Neural Network Robust Model Efficient Architecture Depth Wise Convolution Efficient CNN Bolt Rotation

April 5, 2022

SD-Conv: Towards the Parameter-Efficiency of Dynamic Convolution
Shwai He, Chenbo Jiang, Daize Dong, Liang Ding
Parameter Efficient Dynamic Convolution Traditional Convolution Efficient CNN Pruning Mask

March 13, 2022

AugShuffleNet: Communicate More, Compute Less
Longqing Ye
Timely Communication Shuffle Model Efficient CNN Bottleneck Structure

March 1, 2022

Low-Cost On-device Partial Domain Adaptation (LoCO-PDA): Enabling efficient CNN retraining on edge devices
Aditya Rajagopal, Christos-Savvas Bouganis
Convolutional Neural Network Domain Adaptation Efficient CNN Training Data Distribution Partial Domain Adaptation

January 9, 2022

ThreshNet: An Efficient DenseNet Using Threshold Mechanism to Reduce Connections
Rui-Yang Ju, Ting-Yu Lin, Jia-Hao Jian, Jen-Shiun Chiang, Wei-Bin Yang
Model Compression Neural Network Architecture New Connection SE SPP DenseNet Different Convolutional Neural Network EfficientNet Algorithm Efficient CNN