# ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blocks

Source: [ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blocks](<https://www.sota2.com/research/paper/68fa66401e209131cea0f334>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-28T13:56:40.082Z

Authors: Xiaohan Ding, Yuchen Guo, Guiguang Ding, Jungong Han

Published: 2019-08-11T16:06:58.000Z

## Abstract

As designing appropriate Convolutional Neural Network (CNN) architecture in the context of a given application usually involves heavy human works or numerous GPU hours, the research community is soliciting the architecture-neutral CNN structures, which can be easily plugged into multiple mature architectures to improve the performance on our real-world applications. We propose Asymmetric Convolution Block (ACB), an architecture-neutral structure as a CNN building block, which uses 1D asymmetric convolutions to strengthen the square convolution kernels. For an off-the-shelf architecture, we replace the standard square-kernel convolutional layers with ACBs to construct an Asymmetric Convolutional Network (ACNet), which can be trained to reach a higher level of accuracy. After training, we equivalently convert the ACNet into the same original architecture, thus requiring no extra computations anymore. We have observed that ACNet can improve the performance of various models on CIFAR and ImageNet by a clear margin. Through further experiments, we attribute the effectiveness of ACB to its capability of enhancing the model's robustness to rotational distortions and strengthening the central skeleton parts of square convolution kernels.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Semantic Segmentation on PASCAL-Context 60 classes (test)](<https://www.sota2.com/research/sota/semantic-segmentation-on-pascal-context-60-classes-test.md>): Semantic segmentation; PASCAL-Context 60 classes (test); mIoU: 54.1.

- [Image Classification on ImageNet (val) (Top-1/Top-5 Accuracy)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-val-top-1-top-5-accuracy.md>): Image Classification; ImageNet (val); Top-1 Accuracy: 72.1.

- [Semantic Segmentation on ADE20K (val)](<https://www.sota2.com/research/sota/semantic-segmentation-on-ade20k-val.md>): Semantic segmentation; ADE20K (val); mIoU: 45.9.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/1908.03930>)

[PDF](<https://arxiv.org/pdf/1908.03930.pdf>)
