English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We propose RepMLP, a multi-layer-perceptron-style neural network buildingblock 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, butworse at capturing the local structures, hence usually less favored for imagerecognition. We propose a structural re-parameterization technique that addslocal prior into an FC to make it powerful for image recognition. Specifically,we construct convolutional layers inside a RepMLP during training and mergethem into the FC for inference. On CIFAR, a simple pure-MLP model showsperformance very close to CNN. By inserting RepMLP in traditional CNN, weimprove ResNets by 1.8% accuracy on ImageNet, 2.9% for face recognition, and2.3% mIoU on Cityscapes with lower FLOPs. Our intriguing findings highlightthat combining the global representational capacity and positional perceptionof FC with the local prior of convolution can improve the performance of neuralnetwork 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 athttps://github.com/DingXiaoH/RepMLP.

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