AI Paper Atlas

Research map / Continual learning / Class-incremental

R-DFCIL: Relation-Guided Representation Learning for Data-Free Class Incremental Learning

Qiankun Gao, Chen Zhao, Bernard Ghanem, Jian Zhang

ECCV 2022 · Continual learning › Class-incremental · approach: Exemplar-efficient class-incremental learning · 88 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
add relational knowledge distillation (RKD) (ablation)R-DFCIL · CIFAR100 (equal split) N=20last accuracy: 21.63 -> 30.75Improved
add hard knowledge distillation (HKD) on synthetic data (ablation)R-DFCIL · CIFAR100 (equal split) N=20last accuracy: 5.37 -> 30.75Improved
add classification head refinement (CHR) (ablation)R-DFCIL · CIFAR100 (equal split) N=20last accuracy: 23.75 -> 30.75Improved
add adaptive linear transformation functions in RKD (ablation)R-DFCIL · CIFAR100 (equal split) N=20last accuracy: about 2%Improved
replace RKD with HKD on new data (ablation)R-DFCIL · CIFAR100 N=20 (single run)last accuracy A_20: 30.92% -> 21.35%Hurt
replace HKD with RKD on synthetic old data (ablation)R-DFCIL · CIFAR100 N=20 (single run)last accuracy A_20: 30.92% -> 5.81%Hurt
use RKD instead of newer relational KD (CRCD, CRD) (ablation)R-DFCIL · CIFAR100average accuracy: 5% (CRCD) and 10% (CRD)Improved
raise hard KD weight lambda_hkd to 5.0 (ablation)R-DFCIL · CIFAR100 N=5accuracyHurt
vary lambda_rkd over 0.05, 0.5, 5.0 (ablation)R-DFCIL · CIFAR100 N=5accuracyNo clear effect
use R-DFCIL instead of ABD with more tasks (N=5 to N=20)ABD · CIFAR100 (equal split)last accuracy: +3.11 at N=5, +8.46 at N=20Improved
use R-DFCIL instead of ABDABD · Tiny-ImageNet200 (equal split) N=20last accuracy: 15.20 -> 24.43Improved
replace real old data with synthetic old data in replay methodsUCIR/PODNet · CIFAR100 (50-class first task) N=26last accuracy: 50.08 -> 9.62 (UCIR); 49.14 -> 20.18 (PODNet)Hurt

Later papers that compared with this method

4 later accepted papers report R-DFCIL as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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