AI Paper Atlas

Research map / Continual learning / Online and replay-based

CRAM: Large Scale Video Continual Learning with Bootstrapped Compression

Shivani Mall, Joao F. Henriques

ICCV 2025 · Continual learning › Online and replay-based · approach: Plug-and-play online training enhancements · 1 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 code refresh with drift-free (no refresh) baseline, i.e. DF vs CRAM uses 13779 Mb vs 1404 Mb memory (ablation)CRAM · Kinetics-700 (incremental setting)eval accuracy (%): 37.5 vs 38.80No clear effect
add code refreshing on top of compressed vision (CV) (ablation)CRAM · Kinetics-700 (incremental setting)eval accuracy (%): 26.5 -> 38.80Improved
add code refreshing on top of compressed vision (CV) (ablation)CRAM · Kinetics-700 (incremental setting)average forgetting (%): 60.4 -> 15.20Improved
add rehearsal buffer of compressed codes vs naive SGD (ablation)CRAM · Kinetics-700 (incremental setting)eval accuracy (%): 10.1 -> 38.80Improved
use 50x compression instead of 256x (ablation)CRAM · Kinetics-700 (incremental setting)eval accuracy (%): 38.80 -> 40.1Improved
use 50x compression instead of 256x (ablation)CRAM · Kinetics-700 (pretraining setting)eval accuracy (%): 46.50 -> 47.4Improved
train 40 epochs per task instead of 30 (ablation)CRAM · Kinetics-700 (incremental setting)eval accuracy (%): 38.80 -> 41.9Improved
train 40 epochs per task instead of 30 (ablation)CRAM · EpicKitchens-100 (incremental setting)eval accuracy (%): 28.10 -> 31.1Improved
use 15 classes per task for 45 tasks instead of 20 per task (ablation)CRAM · Kinetics-700 (incremental setting)eval accuracy (%): 38.80 -> 36.1Hurt
use 20 classes per task for 15 tasks instead of 30 per task (ablation)CRAM · Kinetics-700 (pretraining setting)eval accuracy (%): 46.50 -> 47.0No clear effect
replace REMIND with CRAMREMIND · Kinetics-700 (pretraining setting)eval accuracy (%): 38.20 -> 46.50Improved
replace SMILE with CRAMSMILE · Kinetics-700 (incremental setting)eval accuracy (%): 29.20 -> 38.80Improved

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