Log-Polar Space Convolution for Convolutional Neural Networks
This addresses the problem of small local receptive fields in convolutional neural networks for computer vision researchers, representing an incremental improvement with a novel method for a known bottleneck.
The paper tackles the limited receptive fields in lower layers of CNNs due to small convolution kernels by proposing a log-polar space convolution (LPSC) method, which uses elliptical kernels to exponentially increase the receptive field while maintaining parameter count, achieving effectiveness in experiments on various tasks and datasets.
Convolutional neural networks use regular quadrilateral convolution kernels to extract features. Since the number of parameters increases quadratically with the size of the convolution kernel, many popular models use small convolution kernels, resulting in small local receptive fields in lower layers. This paper proposes a novel log-polar space convolution (LPSC) method, where the convolution kernel is elliptical and adaptively divides its local receptive field into different regions according to the relative directions and logarithmic distances. The local receptive field grows exponentially with the number of distance levels. Therefore, the proposed LPSC not only naturally encodes local spatial structures, but also greatly increases the single-layer receptive field while maintaining the number of parameters. We show that LPSC can be implemented with conventional convolution via log-polar space pooling and can be applied in any network architecture to replace conventional convolutions. Experiments on different tasks and datasets demonstrate the effectiveness of the proposed LPSC. Code is available at https://github.com/BingSu12/Log-Polar-Space-Convolution.