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

This paper studies the problem of designing compact binary architectures forvision multi-layer perceptrons (MLPs). We provide extensive analysis on thedifficulty of binarizing vision MLPs and find that previous binarizationmethods perform poorly due to limited capacity of binary MLPs. In contrast withthe traditional CNNs that utilizing convolutional operations with large kernelsize, fully-connected (FC) layers in MLPs can be treated as convolutionallayers with kernel size $1\times1$. Thus, the representation ability of the FClayers will be limited when being binarized, and places restrictions on thecapability of spatial mixing and channel mixing on the intermediate features.To this end, we propose to improve the performance of binary MLP (BiMLP) modelby enriching the representation ability of binary FC layers. We design a novelbinary block that contains multiple branches to merge a series of outputs fromthe same stage, and also a universal shortcut connection that encourages theinformation flow from the previous stage. The downsampling layers are alsocarefully designed to reduce the computational complexity while maintaining theclassification performance. Experimental results on benchmark datasetImageNet-1k demonstrate the effectiveness of the proposed BiMLP models, whichachieve state-of-the-art accuracy compared to prior binary CNNs. The MindSporecode is available at\url{https://gitee.com/mindspore/models/tree/master/research/cv/BiMLP}.

BiMLP: Compact Binary Architectures for Vision Multi-Layer Perceptrons