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

Continual learning aims to learn a sequence of tasks by leveraging theknowledge acquired in the past in an online-learning manner while being able toperform well on all previous tasks, this ability is crucial to the artificialintelligence (AI) system, hence continual learning is more suitable for mostreal-word and complex applicative scenarios compared to the traditionallearning pattern. However, the current models usually learn a genericrepresentation base on the class label on each task and an effective strategyis selected to avoid catastrophic forgetting. We postulate that selecting therelated and useful parts only from the knowledge obtained to perform each taskis more effective than utilizing the whole knowledge. Based on this fact, inthis paper we propose a new framework, named Selecting Related Knowledge forOnline Continual Learning (SRKOCL), which incorporates an additional efficientchannel attention mechanism to pick the particular related knowledge for everytask. Our model also combines experience replay and knowledge distillation tocircumvent the catastrophic forgetting. Finally, extensive experiments areconducted on different benchmarks and the competitive experimental resultsdemonstrate that our proposed SRKOCL is a promised approach against thestate-of-the-art.

Selecting Related Knowledge via Efficient Channel Attention for Online Continual Learning