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

In the past decade, convolutional neural networks (CNNs) have shownprominence for semantic segmentation. Although CNN models have very impressiveperformance, the ability to capture global representation is stillinsufficient, which results in suboptimal results. Recently, Transformerachieved huge success in NLP tasks, demonstrating its advantages in modelinglong-range dependency. Recently, Transformer has also attracted tremendousattention from computer vision researchers who reformulate the image processingtasks as a sequence-to-sequence prediction but resulted in deteriorating localfeature details. In this work, we propose a lightweight real-time semanticsegmentation network called LETNet. LETNet combines a U-shaped CNN withTransformer effectively in a capsule embedding style to compensate forrespective deficiencies. Meanwhile, the elaborately designed LightweightDilated Bottleneck (LDB) module and Feature Enhancement (FE) module cultivate apositive impact on training from scratch simultaneously. Extensive experimentsperformed on challenging datasets demonstrate that LETNet achieves superiorperformances in accuracy and efficiency balance. Specifically, It only contains0.95M parameters and 13.6G FLOPs but yields 72.8\% mIoU at 120 FPS on theCityscapes test set and 70.5\% mIoU at 250 FPS on the CamVid test dataset usinga single RTX 3090 GPU. The source code will be available athttps://github.com/IVIPLab/LETNet.

Lightweight Real-time Semantic Segmentation Network with Efficient Transformer and CNN