Deep competitive pathway networks

Jia Ren Chang, Yong Sheng Chen

研究成果: Conference article同行評審

摘要

In the design of deep neural architectures, recent studies have demonstrated the benefits of grouping subnetworks into a larger network. For examples, the Inception architecture integrates multi-scale subnetworks and the residual network can be regarded that a residual unit combines a residual subnetwork with an identity shortcut. In this work, we embrace this observation and propose the Competitive Pathway Network (CoPaNet). The CoPaNet comprises a stack of competitive pathway units and each unit contains multiple parallel residual-type subnetworks followed by a max operation for feature competition. This mechanism enhances the model capability by learning a variety of features in subnetworks. The proposed strategy explicitly shows that the features propagate through pathways in various routing patterns, which is referred to as pathway encoding of category information. Moreover, the cross-block shortcut can be added to the CoPaNet to encourage feature reuse. We evaluated the proposed CoPaNet on four object recognition benchmarks: CIFAR-10, CIFAR-100, SVHN, and ImageNet. CoPaNet obtained the state-of-the-art or comparable results using similar amounts of parameters. The code of CoPaNet is available at: https://github.com/JiaRenChang/CoPaNet.

原文English
頁(從 - 到)264-278
頁數15
期刊Journal of Machine Learning Research
77
出版狀態Published - 9月 2017
事件9th Asian Conference on Machine Learning, ACML 2017 - Seoul, Korea, Republic of
持續時間: 15 11月 201717 11月 2017

指紋

深入研究「Deep competitive pathway networks」主題。共同形成了獨特的指紋。

引用此