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

Self-supervised learning (SSL), especially contrastive methods, has raisedattraction recently as it learns effective transferable representations withoutsemantic annotations. A common practice for self-supervised pre-training is touse as much data as possible. For a specific downstream task, however,involving irrelevant data in pre-training may degenerate the downstreamperformance, observed from our extensive experiments. On the other hand, forexisting SSL methods, it is burdensome and infeasible to use differentdownstream-task-customized datasets in pre-training for different tasks. Toaddress this issue, we propose a novel SSL paradigm called Scalable DynamicRouting (SDR), which can be trained once and deployed efficiently to differentdownstream tasks with task-customized pre-trained models. Specifically, weconstruct the SDRnet with various sub-nets and train each sub-net with only onesubset of the data by data-aware progressive training. When a downstream taskarrives, we route among all the pre-trained sub-nets to get the best along withits corresponding weights. Experiment results show that our SDR can train 256sub-nets on ImageNet simultaneously, which provides better transfer performancethan a unified model trained on the full ImageNet, achieving state-of-the-art(SOTA) averaged accuracy over 11 downstream classification tasks and AP onPASCAL VOC detection task.

Task-Customized Self-Supervised Pre-training with Scalable Dynamic Routing