RepMLP: Re-parameterizing Convolutions into Fully-connected Layers for Image Recognition

TL;DR

RepMLP introduces a re-parameterization technique that combines the benefits of fully-connected layers and convolutional layers, enhancing image recognition performance while maintaining efficiency.

Contribution

The paper presents RepMLP, a novel neural network block that merges convolutional local priors with fully-connected layers through structural re-parameterization, improving accuracy and speed.

Findings

Pure-MLP model on CIFAR rivals CNN performance.

Inserting RepMLP improves ResNet accuracy on ImageNet.

Enhances face recognition and semantic segmentation with lower FLOPs.

Abstract

We propose RepMLP, a multi-layer-perceptron-style neural network building block for image recognition, which is composed of a series of fully-connected (FC) layers. Compared to convolutional layers, FC layers are more efficient, better at modeling the long-range dependencies and positional patterns, but worse at capturing the local structures, hence usually less favored for image recognition. We propose a structural re-parameterization technique that adds local prior into an FC to make it powerful for image recognition. Specifically, we construct convolutional layers inside a RepMLP during training and merge them into the FC for inference. On CIFAR, a simple pure-MLP model shows performance very close to CNN. By inserting RepMLP in traditional CNN, we improve ResNets by 1.8% accuracy on ImageNet, 2.9% for face recognition, and 2.3% mIoU on Cityscapes with lower FLOPs. Our intriguing findings highlight that combining the global representational capacity and positional perception of FC with the local prior of convolution can improve the performance of neural network with faster speed on both the tasks with translation invariance (e.g., semantic segmentation) and those with aligned images and positional patterns (e.g., face recognition). The code and models are available at https://github.com/DingXiaoH/RepMLP.