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

Since being introduced in 2020, Vision Transformers (ViT) has been steadilybreaking the record for many vision tasks and are often described as``all-you-need" to replace ConvNet. Despite that, ViTs are generallycomputational, memory-consuming, and unfriendly for embedded devices. Inaddition, recent research shows that standard ConvNet if redesigned and trainedappropriately can compete favorably with ViT in terms of accuracy andscalability. In this paper, we adopt the modernized structure of ConvNet todesign a new backbone for action recognition. Particularly, our main target isto serve for industrial product deployment, such as FPGA boards in which onlystandard operations are supported. Therefore, our network simply consists of 2Dconvolutions, without using any 3D convolution, long-range attention plugin, orTransformer blocks. While being trained with much fewer epochs (5x-10x), ourbackbone surpasses the methods using (2+1)D and 3D convolution, and achievecomparable results with ViT on two benchmark datasets.

VidConv: A modernized 2D ConvNet for Efficient Video Recognition