4.6 Review

A Comparison of Pooling Methods for Convolutional Neural Networks

Related references

Note: Only part of the references are listed.
Article Computer Science, Information Systems

Mixed fuzzy pooling in convolutional neural networks for image classification

Teena Sharma et al.

Summary: This paper proposes a mixed fuzzy pooling method that combines max pooling and fuzzy pooling to improve image classification accuracy in the CNN architecture. Experimental results show that the proposed mixed fuzzy pooling outperforms traditional pooling operations.

MULTIMEDIA TOOLS AND APPLICATIONS (2023)

Article Computer Science, Theory & Methods

Avoiding Overfitting: A Survey on Regularization Methods for Convolutional Neural Networks

Claudio Filipi Goncalves Dos Santos et al.

Summary: This paper analyzes the application of various regularization methods developed in recent years to CNN models, including data augmentation, internal changes, and label transformations, demonstrating significant improvements in different CNN models. What sets this paper apart from other surveys is that it focuses on papers published in the last five years and emphasizes the reproducibility of the results.

ACM COMPUTING SURVEYS (2022)

Article Computer Science, Artificial Intelligence

Computer vision-based approach for smart traffic condition assessment at the railroad grade crossing

Feng Guo et al.

Summary: In this study, researchers introduced an automated video analysis, detection and tracking system to assess traffic conditions, analyze behaviors of blocked vehicles at grade crossings, and predict decongestion time under a simplified scenario. A new YOLOv3-SPP+ model was developed to improve detection performance by dividing images into finer to coarser levels and enhancing local features. The results indicated a pattern in vehicle behavior during and after crossing blockages, with a good linear correlation between decongestion time and the number of blocked vehicles.

ADVANCED ENGINEERING INFORMATICS (2022)

Article Engineering, Mechanical

An image-level weakly supervised segmentation method for No-service rail surface defect with size prior

Defu Zhang et al.

Summary: A novel image-level weakly supervised segmentation formulation is proposed for surface defect segmentation of no-service rail, decomposing defects into sub-categories based on size prior information and utilizing a pooling combination module to make full use of size attributes. Experimental results show that the method is effective and outperforms the state-of-the-art methods.

MECHANICAL SYSTEMS AND SIGNAL PROCESSING (2022)

Article Engineering, Chemical

Prediction of ball milling performance by a convolutional neural network model and transfer learning

Yaoyu Li et al.

Summary: This study proposes a three-phase modelling framework using the CNN method to predict the performance of a ball mill based on externally measurable variables. The model shows accurate predictions of particle size distributions and grinding rates in the pre-training and training phases. Application of the model in a large mill system demonstrates its robust performance with limited training datasets. The study also highlights the advantages of transfer learning and discusses the potentials and limitations of the model.

POWDER TECHNOLOGY (2022)

Article Computer Science, Information Systems

VISPNN: VGG-Inspired Stochastic Pooling Neural Network

Shui-Hua Wang et al.

Summary: This study presents a novel artificial intelligence model, VISPNN, for more accurate recognition of alcoholism, which outperforms internal networks and state-of-the-art methods.

CMC-COMPUTERS MATERIALS & CONTINUA (2022)

Article Computer Science, Software Engineering

CNN Based Features Extraction for Age Estimation and Gender Classification

Mohammed Kamel Benkaddour

Summary: The study focuses on automated predictions of age and gender through deep convolutional neural networks, showing significant improvements in accuracy and robustness compared to conventional methods. The proposed approach outperforms traditional methods, achieving state-of-the-art performance in age and gender scoring.

INFORMATICA-AN INTERNATIONAL JOURNAL OF COMPUTING AND INFORMATICS (2021)

Article Computer Science, Artificial Intelligence

Combining max-pooling and wavelet pooling strategies for semantic image segmentation

Andre de Souza Brito et al.

Summary: This paper presents a novel multi-pooling architecture generated by combining the advantages of wavelet and max-pooling operations in convolutional neural networks (CNNs) for semantic segmentation tasks. Experimental results show that this multi-pooling architecture can enhance the performance of aerial image segmentation tasks, achieving results comparable to state-of-the-art approaches.

EXPERT SYSTEMS WITH APPLICATIONS (2021)

Article Computer Science, Information Systems

Multi-class brain tumor classification using residual network and global average pooling

R. Lokesh Kumar et al.

Summary: This study introduces a new model using deep neural network for brain tumor detection and diagnosis, addressing the issues of vanishing gradient and overfitting. Through simulation evaluation with a dataset of 3064 images, the proposed model demonstrates superior classification accuracy.

MULTIMEDIA TOOLS AND APPLICATIONS (2021)

Article Engineering, Civil

Rapid Detection of Blind Roads and Crosswalks by Using a Lightweight Semantic Segmentation Network

Zhengcai Cao et al.

Summary: A lightweight semantic segmentation network is proposed to quickly and accurately segment blind roads and crosswalks in a complex road environment. The proposed method uses different modules to ensure segmentation accuracy and has been verified to be effective by collecting and producing a dataset from a real environment.

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS (2021)

Article Agriculture, Multidisciplinary

A high-precision detection method of hydroponic lettuce seedlings status based on improved Faster RCNN

Zhenbo Li et al.

Summary: This study proposes an automatic detection method for hydroponic lettuce seedlings based on an improved Faster RCNN framework to improve the efficiency and reduce the cost of seedlings sorting in the raising process. Using HRNet for image feature extraction and focal loss along with RoI Align to enhance the accuracy of seedlings detection, the method achieves a high mean average precision of 86.2% for hydroponic lettuce seedlings, outperforming other detectors like RetinaNet and SSD.

COMPUTERS AND ELECTRONICS IN AGRICULTURE (2021)

Article Computer Science, Artificial Intelligence

Hybrid MPSO-CNN: Multi-level Particle Swarm optimized hyperparameters of Convolutional Neural Network

Pratibha Singh et al.

Summary: Recent advances in swarm inspired optimization algorithms have shown extensive acceptance in solving a wide range of real-world problems. The proposed Multi-level Particle Swarm Optimization (MPSO) algorithm allows for finding the architecture and hyperparameters of Convolutional Neural Networks simultaneously, reducing the burden of human experts. Using multiple swarms at different levels and a sigmoid-like inertia weight, the algorithm effectively explores the search space and avoids premature convergence to local optima. The experimental results demonstrate the effectiveness of PSO in learning deep neural architectures across various benchmark datasets.

SWARM AND EVOLUTIONARY COMPUTATION (2021)

Article Mathematics, Interdisciplinary Applications

A novel stochastic resonance model based on bistable stochastic pooling network and its application

Wenyue Zhang et al.

Summary: The study presents a BSPN model based on the traditional SR model to improve the efficiency of weak fault diagnosis, optimizing optimal weight vector through linear weighted optimization on random noise-optimized BSPN output vector, and validating the efficacy of the BSPN system through bearing data collected by two different experimental systems.

CHAOS SOLITONS & FRACTALS (2021)

Article Computer Science, Artificial Intelligence

Mixed pooling and richer attention feature fusion for crack detection

Qiang Zhou et al.

Summary: This study introduces a novel network architecture for crack detection, incorporating richer feature fusion, attention mechanisms, and mixed pooling modules. By utilizing spatial attention at low-level feature maps and channel-wise attention at higher levels, in addition to richer convolutional features, a more accurate crack prediction is achieved.

PATTERN RECOGNITION LETTERS (2021)

Article Computer Science, Hardware & Architecture

Deep Learning for AI

Yoshua Bengio et al.

Summary: Research on artificial neural networks is motivated by the observation that human intelligence emerges from parallel networks of simple non-linear neurons, leading to the question of how these networks can learn complicated internal representations.

COMMUNICATIONS OF THE ACM (2021)

Article Computer Science, Hardware & Architecture

Machine-Learning-Aided Optical Fiber Communication System

Xiaolong Pan et al.

Summary: The article discusses the application of machine learning in optical communication network systems, presenting three use cases to evaluate the feasibility of the proposed architecture. Machine learning algorithms are successfully applied in the transmission and network layers to enhance performance and address challenges in optical communications, showing promising results through extensive simulations and experiments. This research aims to stimulate further exploration and research in the field of machine learning in optical communications.

IEEE NETWORK (2021)

Article Computer Science, Artificial Intelligence

A comparative analysis of pooling strategies for convolutional neural network based Hindi ASR

Vishal Passricha et al.

JOURNAL OF AMBIENT INTELLIGENCE AND HUMANIZED COMPUTING (2020)

Review Computer Science, Artificial Intelligence

Interpretation of intelligence in CNN-pooling processes: a methodological survey

Nadeem Akhtar et al.

NEURAL COMPUTING & APPLICATIONS (2020)

Article Computer Science, Artificial Intelligence

Effect of pooling strategy on convolutional neural network for classification of hyperspectral remote sensing images

Somenath Bera et al.

IET IMAGE PROCESSING (2020)

Article Computer Science, Artificial Intelligence

No-reference mesh visual quality assessment via ensemble of convolutional neural networks and compact multi-linear pooling

Ilyass Abouelaziz et al.

PATTERN RECOGNITION (2020)

Article Computer Science, Artificial Intelligence

Automated diagnosis of multi-class brain abnormalities using MRI images: A deep convolutional neural network based method

Deepak Ranjan Nayak et al.

PATTERN RECOGNITION LETTERS (2020)

Article Plant Sciences

Linking functional traits to multiscale statistics of leaf venation networks

Benjamin Blonder et al.

NEW PHYTOLOGIST (2020)

Article Computer Science, Artificial Intelligence

Deep neural network to extract high-level features and labels in multi-label classification problems

Marilyn Bello et al.

NEUROCOMPUTING (2020)

Article Computer Science, Artificial Intelligence

Ensemble transfer CNNs driven by multi-channel signals for fault diagnosis of rotating machinery cross working conditions

He Zhiyi et al.

KNOWLEDGE-BASED SYSTEMS (2020)

Article Computer Science, Information Systems

Data-Driven Based Tiny-YOLOv3 Method for Front Vehicle Detection Inducing SPP-Net

Xiaolan Wang et al.

IEEE ACCESS (2020)

Article Computer Science, Artificial Intelligence

Overfitting remedy by sparsifying regularization on fully-connected layers of CNNs

Qi Xu et al.

NEUROCOMPUTING (2019)

Article Computer Science, Interdisciplinary Applications

Superpixel based land cover classification of VHR satellite image combining multi-scale CNN and scale parameter estimation

Yangyang Chen et al.

EARTH SCIENCE INFORMATICS (2019)

Article Computer Science, Interdisciplinary Applications

A new approach for arrhythmia classification using deep coded features and LSTM networks

Ozal Yildirim et al.

COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE (2019)

Article Ecology

Deep convolution neural network for image recognition

Boukaye Boubacar Traore et al.

ECOLOGICAL INFORMATICS (2018)

Review Health Care Sciences & Services

Medical Image Analysis using Convolutional Neural Networks: A Review

Syed Muhammad Anwar et al.

JOURNAL OF MEDICAL SYSTEMS (2018)

Proceedings Paper Computer Science, Theory & Methods

InceptB: A CNN Based Classification Approach for Recognizing Traditional Bengali Games

Mohammad Shakirul Islam et al.

8TH INTERNATIONAL CONFERENCE ON ADVANCES IN COMPUTING & COMMUNICATIONS (ICACC-2018) (2018)

Article Acoustics

Improved Audio Scene Classification Based on Label-Tree Embeddings and Convolutional Neural Networks

Huy Phan et al.

IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING (2017)

Article Computer Science, Artificial Intelligence

Rank-based pooling for deep convolutional neural networks

Zenglin Shi et al.

NEURAL NETWORKS (2016)

Article Computer Science, Artificial Intelligence

Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition

Kaiming He et al.

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE (2015)

Article Computer Science, Artificial Intelligence

Towards dropout training for convolutional neural networks

Haibing Wu et al.

NEURAL NETWORKS (2015)