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

As neural networks are increasingly being applied to real-world applications,mechanisms to address distributional shift and sequential task learning withoutforgetting are critical. Methods incorporating network expansion have shownpromise by naturally adding model capacity for learning new tasks whilesimultaneously avoiding catastrophic forgetting. However, the growth in thenumber of additional parameters of many of these types of methods can becomputationally expensive at larger scales, at times prohibitively so. Instead,we propose a simple task-specific feature map transformation strategy forcontinual learning, which we call Efficient Feature Transformations (EFTs).These EFTs provide powerful flexibility for learning new tasks, achieved withminimal parameters added to the base architecture. We further propose a featuredistance maximization strategy, which significantly improves task prediction inclass incremental settings, without needing expensive generative models. Wedemonstrate the efficacy and efficiency of our method with an extensive set ofexperiments in discriminative (CIFAR-100 and ImageNet-1K) and generative (LSUN,CUB-200, Cats) sequences of tasks. Even with low single-digit parameter growthrates, EFTs can outperform many other continual learning methods in a widerange of settings.

Efficient Feature Transformations for Discriminative and Generative Continual Learning