Sheng-Jun Huang

20

Papers

5

Total Citations

Papers (20)

One-shot Active Learning Based on Lewis Weight Sampling for Multiple Deep Models

Learning to Trust Bellman Updates: Selective State-Adaptive Regularization for Offline RL

Rethinking Epistemic and Aleatoric Uncertainty for Active Open-Set Annotation: An Energy-Based Approach

Bidirectional Uncertainty-Based Active Learning for Open-Set Annotation

Dual-Decoupling Learning and Metric-Adaptive Thresholding for Semi-Supervised Multi-Label Learning

StructSR: Refuse Spurious Details in Real-World Image Super-Resolution

Improving Generalization of Deep Neural Networks by Optimum Shifting

MLC-NC: Long-Tailed Multi-Label Image Classification Through the Lens of Neural Collapse

Unlocking the Power of Open Set: A New Perspective for Open-Set Noisy Label Learning

Dirichlet-Based Prediction Calibration for Learning with Noisy Labels

Counterfactual Reasoning for Multi-Label Image Classification via Patching-Based Training

Active Learning for Open-Set Annotation

Improving Lens Flare Removal with General-Purpose Pipeline and Multiple Light Sources Recovery

Multi-Label Knowledge Distillation

Better Safe Than Sorry: Preventing Delusive Adversaries with Adversarial Training

NeurIPS 2021arXiv

Multi-Label Learning with Pairwise Relevance Ordering

Label-Aware Global Consistency for Multi-Label Learning with Single Positive Labels

Can Adversarial Training Be Manipulated By Non-Robust Features?

NeurIPS 2022arXiv

Active Learning for Multiple Target Models

Class-Distribution-Aware Pseudo-Labeling for Semi-Supervised Multi-Label Learning