REVIEW 6 cited by
In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learning
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
The recent research in semi-supervised learning (SSL) is mostly dominated by consistency regularization based methods which achieve strong performance. However, they heavily rely on domain-specific data augmentations, which are not easy to generate for all data modalities. Pseudo-labeling (PL) is a general SSL approach that does not have this constraint but performs relatively poorly in its original formulation. We argue that PL underperforms due to the erroneous high confidence predictions from poorly calibrated models; these predictions generate many incorrect pseudo-labels, leading to noisy training. We propose an uncertainty-aware pseudo-label selection (UPS) framework which improves pseudo labeling accuracy by drastically reducing the amount of noise encountered in the training process. Furthermore, UPS generalizes the pseudo-labeling process, allowing for the creation of negative pseudo-labels; these negative pseudo-labels can be used for multi-label classification as well as negative learning to improve the single-label classification. We achieve strong performance when compared to recent SSL methods on the CIFAR-10 and CIFAR-100 datasets. Also, we demonstrate the versatility of our method on the video dataset UCF-101 and the multi-label dataset Pascal VOC.
Forward citations
Cited by 6 Pith papers
-
CoVar: Confidence-Variance-Guided Pseudo-Label Selection for Semi-Supervised Learning
CoVar selects pseudo-labels by jointly scoring maximum confidence and residual-class variance via SVD-based spectral separation, yielding modest gains on semantic segmentation and classification benchmarks.
-
Learning Semantic-Aware Threshold for Multi-Label Image Recognition with Partial Labels
A class-specific threshold learning method, driven by quantiles of the model's own score distributions plus a ranking loss, improves pseudo-label quality in multi-label recognition with partial labels.
-
OFD-Net: Teacher-Free Reliable Semi-supervised Medical Image Segmentation with Orthogonal Feature Disentanglement Net of Foreground-Background
A single-network semi-supervised segmenter using orthogonal foreground/background feature disentanglement to weight pseudo-labels reports state-of-the-art Dice on four medical benchmarks at 5-20% label rates.
-
Virtual Category-Guided Continual Generalized Category Discovery
Virtual categories for uncertain unlabeled samples plus expanded-neighborhood contrastive learning improve continual generalized category discovery over strong baselines.
-
Label-Free Concept Drift Assessment for Reliable AI in Emerging Wireless Applications
CFPT and TabAutoDrift detect concept drift by comparing macro-F1 scores from pseudo-label or transfer-learning retraining, reaching F1 of 0.94 in fingerprinting and 1.00 in link anomalies without post-deployment labels.
-
RAUM-Net: Regional Attention and Uncertainty-aware Mamba Network
RAUM-Net combines Mamba features, region attention, and MC-dropout uncertainty filtering to improve semi-supervised fine-grained classification under occlusion and label scarcity.
Discussion (0). Sign in to comment.