AI Paper Atlas

Research map / Continual learning / Online and replay-based

KSM: Fast Multiple Task Adaption via Kernel-Wise Soft Mask Learning

Li Yang, Zhezhi He, Junshan Zhang, Deliang Fan

CVPR 2021 · Continual learning › Online and replay-based · approach: Gradient projection and plasticity preservation · 11 citations (OpenAlex)

Check 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
use KSM instead of CPGcontinual learning on Twenty Tasks of CIFAR-100 · Twenty Tasks of CIFAR-100average accuracy: 78.7 -> 79.2Improved
use KSM instead of CPGcontinual learning on Twenty Tasks of CIFAR-100 · Twenty Tasks of CIFAR-100training time (s): 2046 -> 94.3Improved
use KSM instead of the strongest reported baselineImageNet-to-Sketch continual learning · CUBSaccuracy: 83.59 -> 83.81Improved
use KSM instead of the strongest reported baselineImageNet-to-Sketch continual learning · Carsaccuracy: 92.80 -> 92.14Hurt
use KSM instead of the strongest reported baselineImageNet-to-Sketch continual learning · Flowersaccuracy: 96.6 -> 96.94Improved
use KSM instead of the strongest reported baselineImageNet-to-Sketch continual learning · WikiArtaccuracy: 77.15 -> 75.25Hurt
use KSM instead of the strongest reported baselineImageNet-to-Sketch continual learning · Sketchaccuracy: 80.78 -> 81.12Improved
add scaling factors to the weights zeroed by the binary mask (ablation)Piggyback on ImageNet-to-Sketch · ImageNet-to-Sketch (CUBS, Cars, Flowers, WikiArt, Sketch)accuracy: 81.59 -> 82.26; 89.62 -> 91.17; 94.77 -> 95.85; 71.33 -> 73.12; 79.91 -> 80.22Improved
use the softmax trick instead of Piggyback's binary-mask learning (ablation)mask learning on ImageNet-to-Sketch · ImageNet-to-Sketch (CUBS, Cars, Flowers, WikiArt, Sketch)accuracy: 81.59 -> 82.86; 89.62 -> 91.71; 94.77 -> 96.67; 71.33 -> 74.06; 79.91 -> 80.70Improved
use kernel-wise rather than element-wise masks (ablation)Piggyback on ImageNet-to-Sketch · ImageNet-to-Sketch (CUBS, Cars, Flowers, WikiArt, Sketch)accuracy: 81.59 -> 81.76; 89.62 -> 89.57; 94.77 -> 94.88; 71.33 -> 70.30; 79.91 -> 79.95Mixed
use element-wise rather than kernel-wise masks (ablation)full KSM on ImageNet-to-Sketch · ImageNet-to-Sketch (CUBS, Cars, Flowers, WikiArt, Sketch)accuracy: 83.81 -> 83.79; 92.14 -> 92.18; 96.94 -> 96.90; 75.25 -> 75.0; 81.12 -> 81.10Mixed
vary the initial task (ablation)KSM on Twenty Tasks of CIFAR-100 · Twenty Tasks of CIFAR-100average accuracy: 79.2; 78.8; 79.3; 78.0; 78.2Mixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 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