AI Paper Atlas

Research map / Continual learning / Class-incremental

IL2M: Class Incremental Learning With Dual Memory

Eden Belouadah, Adrian Popescu

ICCV 2019 · Continual learning › Class-incremental · approach: Exemplar-efficient class-incremental learning · 315 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 score rectification with class-level means (IL2M 1) (ablation)FT · ILSVRC, Z=10, K=5ktop-1 average accuracy: 32.23 -> 42.20Improved
add score rectification with class-level means (IL2M 1) (ablation)FT · ILSVRC, Z=10, K=20ktop-1 average accuracy: 51.13 -> 53.45Improved
add model-level mean normalization alone (IL2M 2) (ablation)FT · ILSVRC, Z=10, K=5ktop-1 average accuracy: 32.23 -> 31.74Hurt
add model-level mean normalization alone (IL2M 2) (ablation)FT · ILSVRC, Z=10, K=20ktop-1 average accuracy: 51.13 -> 51.94Improved
add model-level means on top of class-level means (IL2M 1+2) (ablation)IL2M 1 · ILSVRC, Z=10, K=10ktop-1 average accuracy: 47.64 -> 49.57Improved
add restriction to rectify only images first predicted as new class (ablation)IL2M 1+2 · ILSVRC, Z=10, K=10ktop-1 average accuracy: 49.57 -> 50.82Improved
ablate distillation loss from iCaRL (FT NEM) (ablation)iCaRL · ILSVRC, Z=10, K=10ktop-1 average accuracy: 33.6 -> 49.1Improved
ablate distillation loss from iCaRL (FT NEM) (ablation)iCaRL · VGGFace2, Z=10, K=10ktop-1 average accuracy: 65.3 -> 87.6Improved
add distillation loss to fine tuning with no memory (ablation)FT · ILSVRC, Z=10, K=0top-1 average accuracy: 18.3 -> 20.8Improved
use NEM classifier instead of network classification layer (FT NEM)FT · Landmarks, Z=10, K=10ktop-1 average accuracy: 90.1 -> 88.5Hurt
add balanced fine-tuning step (FT BAL)FT · ILSVRC, Z=10, K=10ktop-1 average accuracy: 42.3 -> 47.0Improved
add IL2M score rectificationFT · ILSVRC, Z=10, K=20k and K=5ktop-1 average accuracy: +5.3 and +11.9 pointsImproved

Later papers that compared with this method

2 later accepted papers report il2m as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

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