AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

DICE: Diversity in Deep Ensembles via Conditional Redundancy Adversarial Estimation

Alexandre Rame, Matthieu Cord

ICLR 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 6 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
Add conditional redundancy loss to CEB (ablation)DICE ensemble · CIFAR-100, ResNet-32 4-branchTop-1 accuracy (%): 76.98 -> 77.51Improved
Use conditional rather than non-conditional redundancy (ablation)VCEB ensemble · CIFAR-100, ResNet-32 4-branchTop-1 accuracy (%): 77.30 -> 77.51Improved
Use VCEB instead of VIB (ablation)Ensemble without redundancy · CIFAR-100, ResNet-32 4-branchTop-1 accuracy (%): 76.93 -> 76.98No clear effect
Add non-conditional redundancy to VIB (ablation)IBR ensemble · CIFAR-100, ResNet-32 4-branchTop-1 accuracy (%): 76.93 -> 77.25Improved
Add conditional redundancy loss (ablation)DICE ensemble · CIFAR-100, ResNet-110 4-branchTop-1 accuracy (%): 81.17 -> 81.93Improved
Add conditional redundancy loss (ablation)DICE ensemble · CIFAR-100, ResNet-32 4-branchNLL (10^-1), after temperature scaling: 8.11 -> 7.98Improved
Add conditional redundancy loss (ablation)DICE ensemble · CIFAR-100, ResNet-32 4-branchECE (10^-2), before temperature scaling: 5.76 -> 4.05Improved
Scale logits with the discriminator's 1-w confidenceDICE×w · CIFAR-100, TinyImageNet (crop) OODAUROC: 74.7 -> 77.2Improved
Increase the diversity coefficient to 0.6 (ablation)DICE · CIFAR-100, 4-branch ResNet-32Individual accuracyHurt
Set the diversity coefficient to -0.025 (ablation)DICE · CIFAR-100, 4-branch ResNet-32Ensemble diversityHurt
Use adversarial redundancy loss without compression (ablation)Redundancy-only ensemble · CIFAR-100Training stabilityHurt

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