ICAF：具有插值一致性和自適應(yīng)篩選的無監(jiān)督持續(xù)學(xué)習(xí)

打開文本圖片集

中圖分類號(hào)：TP18；TP391.41 文獻(xiàn)標(biāo)識(shí)碼：A 文章編號(hào)：2096-4706（2025）11-0049-06

1，2，1，2（1.，；2.，)

ICAF: Unsupervised Continual Learning with Interpolation Consistency and Adaptive Filtering

YE Gencheng12, WANG Xiang1,2 (1.Beijing KeyLabofTraffc Data Analysisand Mining (Beijing Jiaotong University),Beijingl0o044,China; 2.School ofComputer Science and Technology,Beijing Jiaotong University,Beijing 10oo44,China)

Abstract: Unsupervised Continual Learning addresss the challenge of sequentiall learning tasks without supervision information.However,thissetingoftenleadstocatastropicforgeting.Toovercomethisisue,memorybanksareusedtostore samplesfrompastasks.However,storingandreplayingappropriatesamplesisacomplextask,andthesimpleandcommonly usedrandomselectionis noteffective,especiallyforlong-sequence taskswithlarge-scale samples.Tosolve this isue,an Unsupervised ContinualLearning method with Interpolation Consistencyand AdaptiveMemoryFiltering (ICAF)is proposed.It canefectivelyselect high-qualitysamples byusing thefeaturesleamedfromrandomlyaugmentedsamples,adjust thethreshold during training,andachieveinterpolationconsistency throughaspeciallydesigned linear interpolationanddatacombination strategy,thereby preventing insuffcientlearning causedbyrelying solelyoninterpolatedsamplesfrompasttasks.This method has achieved the best results on multiple experimental datasets.

Keywords: Continual Learning; Unsupervised Learning; Representation Learning; Contrastive Learning; Selfsupervised Learning

0 引言

近年來，持續(xù)學(xué)習(xí)在人工智能領(lǐng)域受到了廣泛關(guān)注。（剩余12294字）