AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Dataset Condensation with Gradient Matching

Bo Zhao, Konda Reddy Mopuri, Hakan Bilen

ICLR 2021 · Architectures and efficiency › RNNs and normalization · approach: Efficient network compression and training · 102 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 selected real examples with gradient-matched synthetic imagesConvNet classifiers trained on condensed sets · MNIST, FashionMNIST, SVHN, CIFAR10accuracy: 91.7$\pm$0.5 vs 89.2$\pm$1.6; 70.5$\pm$0.6 vs 67.0$\pm$1.9; 31.2$\pm$1.4 vs 21.0$\pm$1.5; 28.3$\pm$0.5 vs 21.5$\pm$1.2Improved
increase the number of condensed images per classDataset Condensation · MNIST, FashionMNIST, SVHN, CIFAR10accuracy: 91.7$\pm$0.5 -> 98.8$\pm$0.2; 70.5$\pm$0.6 -> 83.6$\pm$0.4; 31.2$\pm$1.4 -> 82.3$\pm$0.3; 28.3$\pm$0.5 -> 53.9$\pm$0.5Improved
use a class-wise gradient-matching distance grouped by layer and output node (ablation)Dataset Condensation · MNIST, 1 image/class, LeNet and ResNetaccuracy: 65.0$\pm$5.1 -> 85.0$\pm$1.7; 68.8$\pm$7.8 -> 89.4$\pm$0.9Improved
use InstanceNorm in the classifier trained on condensed images (ablation)Dataset Condensation · MNIST, 1 image/classaccuracy: 79.0$\pm$2.2 -> 90.7$\pm$0.7Improved
use BatchNorm in the classifier trained on condensed images (ablation)Dataset Condensation · MNIST, 1 image/classaccuracy: 79.0$\pm$2.2 -> 80.8$\pm$2.0Hurt
use Sigmoid instead of ReLU in the classifier (ablation)Dataset Condensation · MNIST, 1 image/classaccuracy: around $5\%$ lowerHurt
use average pooling rather than max pooling (ablation)Dataset Condensation · MNIST, 1 image/classaccuracy: 91.7$\pm$0.5Improved
use pooling rather than no pooling (ablation)Dataset Condensation · MNIST, 1 image/classaccuracyImproved
train the synthetic images with deeper ConvNets (ablation)Dataset Condensation · MNIST, 1 image/classaccuracyImproved
increase generator network width (ablation)Dataset Condensation · MNIST, 1 image/classaccuracyNo clear effect
train the classifier with 128 filters per block (ablation)Dataset Condensation · MNIST, 1 image/classaccuracyImproved
replace gradient-matched synthetic data with Herding-selected examplesContinual learning rehearsal memory · SVHN -> MNIST -> USPS, T3accuracy: 2.3\%Hurt

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