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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

The goal of Few-Shot Continual Learning (FSCL) is to incrementally learnnovel tasks with limited labeled samples and preserve previous capabilitiessimultaneously, while current FSCL methods are all for the class-incrementalpurpose. Moreover, the evaluation of FSCL solutions is only the cumulativeperformance of all encountered tasks, but there is no work on exploring thedomain generalization ability. Domain generalization is a challenging yetpractical task that aims to generalize beyond training domains. In this paper,we set up a Generalized FSCL (GFSCL) protocol involving both class- anddomain-incremental situations together with the domain generalizationassessment. Firstly, two benchmark datasets and protocols are newly arranged,and detailed baselines are provided for this unexplored configuration. We findthat common continual learning methods have poor generalization ability onunseen domains and cannot better cope with the catastrophic forgetting issue incross-incremental tasks. In this way, we further propose a rehearsal-freeframework based on Vision Transformer (ViT) named Contrastive Mixture ofAdapters (CMoA). Due to different optimization targets of class increment anddomain increment, the CMoA contains two parts: (1) For the class-incrementalissue, the Mixture of Adapters (MoA) module is incorporated into ViT, thencosine similarity regularization and the dynamic weighting are designed to makeeach adapter learn specific knowledge and concentrate on particular classes.(2) For the domain-related issues and domain-invariant representation learning,we alleviate the inner-class variation by prototype-calibrated contrastivelearning. The codes and protocols are available athttps://github.com/yawencui/CMoA.

Generalized Few-Shot Continual Learning with Contrastive Mixture of Adapters