What are the Receptive, Effective Receptive, and Projective Fields of Neurons in Convolutional Neural Networks?
This work addresses a fundamental concept for researchers and practitioners in deep learning to better analyze and improve CNN performance, though it is incremental as it clarifies existing calculations.
The paper explains how to calculate receptive, effective receptive, and projective fields for neurons in convolutional neural networks (CNNs), focusing on CNNs but noting applicability to deconvolutional networks. It aims to enhance understanding and optimization of these networks for real-world applications.
In this work, we explain in detail how receptive fields, effective receptive fields, and projective fields of neurons in different layers, convolution or pooling, of a Convolutional Neural Network (CNN) are calculated. While our focus here is on CNNs, the same operations, but in the reverse order, can be used to calculate these quantities for deconvolutional neural networks. These are important concepts, not only for better understanding and analyzing convolutional and deconvolutional networks, but also for optimizing their performance in real-world applications.