AI Paper Atlas

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

CARE: Class-Adaptive Expert Consensus for Reliable Learning with Long-Tailed Noisy Labels

Mengke Li, Haiquan Ling, Lihao Chen, Yang Lu, Yiqun Zhang, Hui Huang

ICML 2026 · Label-efficient and robust learning › Noisy labels · approach: Noisy cross-modal hashing and multi-label · 0 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 CARE class-adaptive expert consensus to CLIP+LACLIP+LA · CIFAR-100-LTN joint noise IF=100 NR=0.5top-1 accuracy (%): 75.3 -> 76.7Improved
add CARE to CLIP+CECLIP+CE · CIFAR-100-LTN symmetric NR=0.6 IF=10top-1 accuracy (%): 74.5 -> 77.2Improved
use all three experts (BE, TE, IE) instead of any single auxiliary expert (ablation)CARE · CIFAR-100-LTN IF=10 joint NR=0.5accuracy (%): 78.7 -> 80.3Improved
use class-wise Kc instead of global K (ablation)CARE · ImageNet-LTNr IF=100 NR=40accuracy (%): 80.7 -> 81.9Improved
use class-wise Kc instead of global K on CIFAR-100-LTN (ablation)CARE · CIFAR-100-LTN IF=10 NR=20 asymmetricaccuracy (%): 80.7 -> 80.8No clear effect
vary form of K (step, exponential, logarithmic, linear) (ablation)CARE · CIFAR-100-LTN IF=100 20% symmetricaccuracy (%): 76.0 -> 77.5Improved
reduce BE weight to 0.5 so a single auxiliary expert can correct labels (ablation)CARE · CIFAR100-IF10-NR50 symmetricnoise rate after correction (%): BE+TE 31.8, BE+IE 29.3 (from 50.0)Improved
CARE versus observed-label expert only: tail-class accuracy (ablation)CARE · CIFAR-100-LTN IF=10 joint NR=0.5tail accuracy (%): 72.3 -> 77.3Improved
CARE versus observed-label expert only: head-class accuracy (ablation)CARE · CIFAR-100-LTN IF=10 joint NR=0.5head accuracy (%): 83.4 -> 81.8Hurt
class-agnostic label correction with lower noise rate but uncorrected tail labelsCLIP fine-tuning · CIFAR-100 IF=10 50% joint noiseaccuracy (%): 79.5 -> 76.8Hurt
class-aware rectification (RLD 3) instead of class-agnosticCLIP fine-tuning · CIFAR-100 IF=10 50% joint noiseaccuracy (%): 79.5 -> 80.9Improved
add CARE on Clothing1Mbaseline · Clothing1Maccuracy (%): 71.2 -> 71.4No clear effect

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