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

It is common practice to reuse models initially trained on different data toincrease downstream task performance. Especially in the computer vision domain,ImageNet-pretrained weights have been successfully used for various tasks. Inthis work, we investigate the impact of transfer learning for segmentationproblems, being pixel-wise classification problems that can be tackled withencoder-decoder architectures. We find that transfer learning the decoder doesnot help downstream segmentation tasks, while transfer learning the encoder istruly beneficial. We demonstrate that pretrained weights for a decoder mayyield faster convergence, but they do not improve the overall model performanceas one can obtain equivalent results with randomly initialized decoders.However, we show that it is more effective to reuse encoder weights trained ona segmentation or reconstruction task than reusing encoder weights trained onclassification tasks. This finding implicates that using ImageNet-pretrainedencoders for downstream segmentation problems is suboptimal. We also propose acontrastive self-supervised approach with multiple self-reconstruction tasks,which provides encoders that are suitable for transfer learning in segmentationproblems in the absence of segmentation labels.

Transfer Learning for Segmentation Problems: Choose the Right Encoder and Skip the Decoder