English

Deep Convolutional Neural Networks with Zero-Padding: Feature Extraction and Learning

Machine Learning 2023-08-01 v1 Computer Vision and Pattern Recognition

Abstract

This paper studies the performance of deep convolutional neural networks (DCNNs) with zero-padding in feature extraction and learning. After verifying the roles of zero-padding in enabling translation-equivalence, and pooling in its translation-invariance driven nature, we show that with similar number of free parameters, any deep fully connected networks (DFCNs) can be represented by DCNNs with zero-padding. This demonstrates that DCNNs with zero-padding is essentially better than DFCNs in feature extraction. Consequently, we derive universal consistency of DCNNs with zero-padding and show its translation-invariance in the learning process. All our theoretical results are verified by numerical experiments including both toy simulations and real-data running.

Keywords

Cite

@article{arxiv.2307.16203,
  title  = {Deep Convolutional Neural Networks with Zero-Padding: Feature Extraction and Learning},
  author = {Zhi Han and Baichen Liu and Shao-Bo Lin and Ding-Xuan Zhou},
  journal= {arXiv preprint arXiv:2307.16203},
  year   = {2023}
}

Comments

18

R2 v1 2026-06-28T11:43:45.757Z