List your company
ResearchFormerly WizWand

Image Classification on CIFAR-10 (test) (Sparsity)

99.6SparsityContinuous Sparsification
10.1633.3856.679.82Dec 10, 2019
Updated 8mo ago

Evaluation results

MethodLinks
Continuous SparsificationBackbone=VGG-16, Accuracy constraint=Within 2% of original dense model2019.1299.6
Network SlimmingBackbone=VGG-16, Accuracy constraint=Within 2% of original dense model2019.1299
GMPBackbone=VGG-16, Accuracy constraint=Within 2% of original dense model2019.1298
Magnitude PruningBackbone=VGG-16, Accuracy constraint=Within 2% of original dense model2019.1297.5
Continuous SparsificationBackbone=ResNet-20, Accuracy constraint=Within 2% of original dense model2019.1294.4
GMPBackbone=ResNet-20, Accuracy constraint=Within 2% of original dense model2019.1286
AMCBackbone=VGG-16, Accuracy constraint=Within 2% of original dense model2019.1286
Network SlimmingBackbone=ResNet-20, Accuracy constraint=Within 2% of original dense model2019.1285
Magnitude PruningBackbone=ResNet-20, Accuracy constraint=Within 2% of original dense model2019.1280
AMCBackbone=ResNet-20, Accuracy constraint=Within 2% of original dense model2019.1250
Louizos et al.Backbone=VGG-16, Accuracy constraint=Within 2% of original dense model2019.1218.2
Louizos et al.Backbone=ResNet-20, Accuracy constraint=Within 2% of original dense model2019.1213.6