AI Paper Atlas

Research map / Continual learning / Class-incremental

Versatile Incremental Learning: Towards Class and Domain-Agnostic Incremental Learning

Min-Yeong Park, Jae-Ho Lee, Gyeong-Moon Park

ECCV 2024 · Continual learning › Class-incremental · approach: Exemplar-efficient class-incremental learning · 9 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace constant threshold 0.5 with a dynamic threshold in the incremental classifier (ablation)ICON · iDigits VILaverage accuracy: 74.79 -> 75.11Improved
place adapters in the first 5 layers instead of the last 5 (ablation)ICON · iDigits VILaverage accuracy: 52.94 -> 75.11Improved
place adapters in all layers instead of the first 5 (ablation)ICON · DomainNet VILaverage accuracy: 53.37 -> 55.14 on DomainNet; 75.11 -> 62.57 on iDigitsMixed
save more than one shift per task in the shift pool (ablation)ICON CAST · iDigits VILaverage accuracy: 75.11 -> 68.45 (10 shifts)Hurt
use more shifts per task on DomainNet (ablation)ICON CAST · DomainNet VILaverage accuracy: 53.37 -> 53.61 (10 shifts)No clear effect
increase number of tasks in the VIL sequence (iDigits 8 to 20 tasks)LAE and ICON · iDigits VILaverage accuracy: ICON 81.59 -> 75.11; LAE 79.67 -> 59.34Hurt
use ICON instead of LAE across numbers of tasksICON · CORe50 VIL 40 tasksaverage accuracy: 77.11 -> 83.18Improved
apply ICON adapters to a ResNet-152 CNN instead of LAEICON · iDigits VIL (ResNet-152)average accuracy: 30.43 -> 37.34Improved
use ICON instead of LAE on larger backbone ViT-L/16ICON · DomainNet VIL (ViT-L/16)average accuracy: 51.80 -> 54.09Improved
use too small gamma in incremental classifier threshold (ablation)ICON · iDigits VILaverage accuracy: gamma 0.5: 72.00 vs gamma 2.0: 75.11Hurt
increase beta (CAST loss weight) on CORe50 (ablation)ICON · CORe50 VILaverage accuracy: 79.85 -> 83.18 (beta 0.01 -> 0.05)Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy