AI Paper Atlas

Research map / Label-efficient and robust learning / Noisy labels

Latent Class-Conditional Noise Model

Jiangchao Yao, Bo Han, Zhihan Zhou, Ya Zhang, Ivor Wai-Hung Tsang

TPAMI 2023 · Label-efficient and robust learning › Noisy labels · approach: Instance-dependent noise modeling · 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
replace S-adaptation noise-transition layer with latent class-conditional modelnoise transition learner · CIFAR-10 symmetric noise r=0.5accuracy (%): 83.4 -> 87.3Improved
use LCCN instead of CEPreAct ResNet-32 · CIFAR-100 asymmetric noise r=0.4accuracy (%): 51.6 -> 65.5Improved
use LCCN instead of DivideMixLCCN · CIFAR-10 asymmetric noise r=0.7accuracy (%): 75.1 -> 79.6Improved
add cross-model training and co-refinement tricks (DivideLCCN)LCCN · CIFAR-10 symmetric noise r=0.5accuracy (%): 87.3 -> 95.3Improved
use DivideLCCN instead of DivideMixDivideMix · CIFAR-10 symmetric noise r=0.9accuracy (%): 75.7 -> 79.2Improved
use DivideLCCN instead of DivideMixDivideMix · CIFAR-100 asymmetric noise r=0.4accuracy (%): 58.3 -> 73.6Improved
add open-set noise handling (LCCN*) (ablation)LCCN · CIFAR-10 open-set noise r=0.7accuracy (%): 69.4 -> 72.4Improved
add a small clean set for calibration (LCCN+) (ablation)LCCN* · CIFAR-10 open-set noise r=0.7accuracy (%): 72.4 -> 87.5Improved
add clean data to LCCN* (LCCN*+) (ablation)LCCN+ · CIFAR-10 open-set noise r=0.7accuracy (%): 87.5 -> 89.4Improved
use DivideLCCN instead of LCCN+ at extreme open-set noiseLCCN+ · CIFAR-10 open-set noise r=0.7accuracy (%): 87.5 -> 79.2Hurt
use LCCN instead of BootstrappingBootstrapping · CIFAR-100 asymmetric noise r=0.5accuracy (%): 34.5 -> 33.7No clear effect
use T-revision instead of CE at high symmetric noiseCE · CIFAR-10 symmetric noise r=0.9accuracy (%): 42.8 -> 28.6Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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