Multiactivation Pooling Method in Convolutional Neural Networks for Image Recognition
Joint Authors
Feng, Wenquan
Zhao, Qi
Lyu, Shuchang
Zhang, Boxue
Source
Wireless Communications and Mobile Computing
Issue
Vol. 2018, Issue 2018 (31 Dec. 2018), pp.1-15, 15 p.
Publisher
Hindawi Publishing Corporation
Publication Date
2018-06-26
Country of Publication
Egypt
No. of Pages
15
Main Subjects
Information Technology and Computer Science
Abstract EN
Convolutional neural networks (CNNs) are becoming more and more popular today.
CNNs now have become a popular feature extractor applying to image processing, big data processing, fog computing, etc.
CNNs usually consist of several basic units like convolutional unit, pooling unit, activation unit, and so on.
In CNNs, conventional pooling methods refer to 2×2 max-pooling and average-pooling, which are applied after the convolutional or ReLU layers.
In this paper, we propose a Multiactivation Pooling (MAP) Method to make the CNNs more accurate on classification tasks without increasing depth and trainable parameters.
We add more convolutional layers before one pooling layer and expand the pooling region to 4×4, 8×8, 16×16, and even larger.
When doing large-scale subsampling, we pick top-k activation, sum up them, and constrain them by a hyperparameter σ.
We pick VGG, ALL-CNN, and DenseNets as our baseline models and evaluate our proposed MAP method on benchmark datasets: CIFAR-10, CIFAR-100, SVHN, and ImageNet.
The classification results are competitive.
American Psychological Association (APA)
Zhao, Qi& Lyu, Shuchang& Zhang, Boxue& Feng, Wenquan. 2018. Multiactivation Pooling Method in Convolutional Neural Networks for Image Recognition. Wireless Communications and Mobile Computing،Vol. 2018, no. 2018, pp.1-15.
https://search.emarefa.net/detail/BIM-1216282
Modern Language Association (MLA)
Zhao, Qi…[et al.]. Multiactivation Pooling Method in Convolutional Neural Networks for Image Recognition. Wireless Communications and Mobile Computing No. 2018 (2018), pp.1-15.
https://search.emarefa.net/detail/BIM-1216282
American Medical Association (AMA)
Zhao, Qi& Lyu, Shuchang& Zhang, Boxue& Feng, Wenquan. Multiactivation Pooling Method in Convolutional Neural Networks for Image Recognition. Wireless Communications and Mobile Computing. 2018. Vol. 2018, no. 2018, pp.1-15.
https://search.emarefa.net/detail/BIM-1216282
Data Type
Journal Articles
Language
English
Notes
Includes bibliographical references
Record ID
BIM-1216282