Image classification in frequency domain with 2SReLU: A second harmonics superposition activation function

Autor(es)

Watanabe, Thomio ^[1] ; Wolf, Denis F. ^[1]

Afiliação do(s) autor(es)

^[1] Univ Sao Paulo, Inst Math & Comp Sci, Sao Paulo - Brazil

• Tipo de documento: Artigo Científico
• Fonte: APPLIED SOFT COMPUTING; v. 112, NOV 2021.
• Citações Web of Science: 0

Resumo

Deep Convolutional Neural Networks are able to identify complex patterns and perform tasks with super-human capabilities. However, besides the exceptional results, they are not completely understood and it is still impractical to hand-engineer similar solutions. In this work, an image classification Convolutional Neural Network and its building blocks are described from a frequency domain perspective. Some network layers have established counterparts in the frequency domain like the convolutional and pooling layers. We propose the 2SReLU layer, a novel non-linear activation function that preserves high frequency components in deep networks. A convolution-free network is presented, and it is demonstrated that in the frequency domain it is possible to achieve competitive results without using the computationally costly convolution operation. A source code implementation in PyTorch is provided at: https://gitlab.com/thomio/2srelu. (C) 2021 Elsevier B.V. All rights reserved. (AU)

• Processo FAPESP: 15/26293-0 - Rastreamento de múltiplos obstáculos em veículos autônomos com fusão de sensores
• Beneficiário: Thomio Watanabe
• Modalidade de apoio: Bolsas no Brasil - Doutorado