Multi-Scale Spatially-Asymmetric Recalibration for Image Classification
This addresses a specific problem in computer vision for image classification, offering an incremental improvement by enhancing contextual feature utilization in existing architectures.
The paper tackled the limitation of spatially-symmetric convolution in image classification by introducing multi-scale spatially-asymmetric recalibration (MS-SAR), which improved performance on CIFAR and ILSVRC2012 tasks with minimal extra parameters and computations.
Convolution is spatially-symmetric, i.e., the visual features are independent of its position in the image, which limits its ability to utilize contextual cues for visual recognition. This paper addresses this issue by introducing a recalibration process, which refers to the surrounding region of each neuron, computes an importance value and multiplies it to the original neural response. Our approach is named multi-scale spatially-asymmetric recalibration (MS-SAR), which extracts visual cues from surrounding regions at multiple scales, and designs a weighting scheme which is asymmetric in the spatial domain. MS-SAR is implemented in an efficient way, so that only small fractions of extra parameters and computations are required. We apply MS-SAR to several popular building blocks, including the residual block and the densely-connected block, and demonstrate its superior performance in both CIFAR and ILSVRC2012 classification tasks.