AI Paper Atlas

Research map / Optimization / Optimizers

Calibrated Stochastic Gradient Descent for Convolutional Neural Networks

Li’an Zhuo, Baochang Zhang, Chen Chen, Qixiang Ye, Jianzhuang Liu, David Doermann

AAAI 2019 · Optimization › Optimizers · approach: SGD momentum and step-size schedules · 9 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
increase K from 1 to 8 (sigma=0.0001) (ablation)CCNN-18 · CIFAR-10error rate (%): 9.48 -> 7.42Improved
increase K from 1 to 4 (sigma=0) (ablation)CCNN-18 · CIFAR-10error rate (%): 9.68 -> 7.82Improved
decrease sigma from 0.001 to 0.0001 (K=4) (ablation)CCNN-18 · CIFAR-10error rate (%): 7.68 -> 7.61Improved
decrease sigma from 0.001 to 0.0001 (K=1) (ablation)CCNN-18 · CIFAR-10error rate (%): 9.60 -> 9.48Improved
add Gaussian noise sigma=0.0001 versus sigma=0 (K=8) (ablation)CCNN-18 · CIFAR-10error rate (%): 7.63 -> 7.42Improved
gradient calibration layers (K=4) instead of Gaussian noise ResNet-18*ResNet-18 · CIFAR-10error rate (%): 9.72 -> 7.61Improved
replace conv layers with GClayers (WRN-28 to CCNNs3-28)WRN · CIFAR-10error rate (%): 4.00 -> 3.81Improved
replace conv layers with GClayers (WRN-28 to CCNNs3-28)WRN · CIFAR-100error rate (%): 19.25 -> 19.11Improved
replace conv layers with GClayers (WRN-40 to CCNNs2-40)WRN · CIFAR-10error rate (%): 4.53 -> 4.62Hurt
replace conv layers with GClayers (WRN-40 to CCNNs2-40)WRN · CIFAR-100error rate (%): 21.18 -> 21.67Hurt
CCNNs-18 versus ResNet-18ResNet-18 · ImageNetTop-1 error: 30.7% -> 29.2%Improved
CCNNs-18 versus ResNet-18ResNet-18 · ImageNetTop-5 error: 10.8% -> 10.3%Improved

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