AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

CODEs: Chamfer Out-of-Distribution Examples Against Overconfidence Issue

Keke Tang, Dingruibo Miao, Weilong Peng, Jianpeng Wu, Yawen Shi, Zhaoquan Gu, Zhihong Tian, Wenping Wang

ICCV 2021 · Label-efficient and robust learning › OOD and anomaly detection · approach: Uncertainty and misclassification detection · 33 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
transform seed examples with Chamfer GANseed OOD examples · CIFAR-10FID to original data: 50.49 -> 36.53Improved
transform seed examples with Chamfer GANseed OOD examples · CIFAR-100FID to original data: 48.83 -> 36.53Improved
transform seed examples with Chamfer GANseed OOD examples · SVHNFID to original data: 83.51 -> 77.94Improved
transform seed examples with Chamfer GANseed OOD examples · CIFAR-10ResNet-18 MMC: 72.53 -> 74.84Improved
train with suppression on CODEsbaseline classifier · CIFAR-10 (train), OOD SVHNOOD MMC (%): 73.38 -> 62.37Improved
train with suppression on CODEsbaseline classifier · CIFAR-10 (train), OOD CIFAR-100OOD MMC (%): 79.47 -> 70.14Improved
train with suppression on CODEsbaseline classifier · CIFAR-10 (train), noiseOOD MMC (%): 69.24 -> 10.27Improved
use Ours++ variant instead of OursCODEs training · CIFAR-10 (train), OOD SVHNOOD MMC (%): 62.37 -> 47.18Improved
train with suppression on CODEsbaseline classifier · CIFAR-10 (train), CIFAR-10 testtest error (%): 5.38 -> 5.58No clear effect
train with suppression on CODEsbaseline classifier · CIFAR-100 (train), CIFAR-100 testtest error (%): 23.34 -> 23.22No clear effect
use auxiliary dataset (OE) instead of CODEsCIFAR-10 classifier · CIFAR-10 (train), CIFAR-10 IDID MMC (%): 97.15 -> 88.87Hurt
train with suppression on CODEsbaseline classifier · CIFAR-10 (train), uniform noiseOOD MMC (%): 99.49 -> 65.30Mixed

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