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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

This paper presents a module, Spatial Cross-scale Convolution (SCSC), whichis verified to be effective in improving both CNNs and Transformers. Nowadays,CNNs and Transformers have been successful in a variety of tasks. Especiallyfor Transformers, increasing works achieve state-of-the-art performance in thecomputer vision community. Therefore, researchers start to explore themechanism of those architectures. Large receptive fields, sparse connections,weight sharing, and dynamic weight have been considered keys to designingeffective base models. However, there are still some issues to be addressed:large dense kernels and self-attention are inefficient, and large receptivefields make it hard to capture local features. Inspired by the above analysesand to solve the mentioned problems, in this paper, we design a general moduletaking in these design keys to enhance both CNNs and Transformers. SCSCintroduces an efficient spatial cross-scale encoder and spatial embed module tocapture assorted features in one layer. On the face recognition task,FaceResNet with SCSC can improve 2.7% with 68% fewer FLOPs and 79% fewerparameters. On the ImageNet classification task, Swin Transformer with SCSC canachieve even better performance with 22% fewer FLOPs, and ResNet with CSCS canimprove 5.3% with similar complexity. Furthermore, a traditional network (e.g.,ResNet) embedded with SCSC can match Swin Transformer's performance.

SCSC: Spatial Cross-scale Convolution Module to Strengthen both CNNs and Transformers