Label 相关论文汇总

1、DLDL

2、MLIC-WCP

3、MLIR-GCN

• Multi-Label Image Recognition with Graph Convolutional Networks, CVPR, 2019

基于图卷积网络的多标签图像识别
旷视科技

https://github.com/nightluo/MLIR-GCN
1
2
3
4

• [14] Semi-Supervised Classification with Graph Convolutional Networks, ICLR, 2017

基于图卷积网络的半监督分类
阿姆斯特丹大学
GCN 开山之作
可扩展的半监督学习方法
基于图上谱卷积的一阶近似
有效应对半监督，并编码图结构和节点特征
https://github.com/nightluo/GCN-py
1
2
3
4
5
6
7

4、C-Tran

5、ACPL

• ACPL: Anti-curriculum Pseudo-labelling for Semi-supervised Medical Image Classification, CVPR, 2022

ACPL：用于半监督医学图像分类的反课程伪标记
澳大利亚机器学习研究所，阿德莱德大学，德国乌尔姆第二大学
反课程伪标签的 SSL 方法
新的伪标记方法 CSDI —— 不需要估计类分类阈值
IM 集合 KNN 分类器，生成更准确的伪标签
ASP 方法保持密度评分在后续训练阶段的准确性
https://github.com/FBLADL/ACPL
1
2
3
4
5
6
7

• [40] When do curricula work, ICLR, 2021

什么时候实施课程学习
谷歌

1
2
3

• Curriculum Learning, ICML,2009

课程学习
新的训练策略，加速收敛至局部 and 全局最优解
1
2

• CurriculumNet: Weakly Supervised Learning from Large-Scale Web Images, ECCV, 2018

课程网络：基于大规模网络图像的弱监督学习
深圳 Malong 人工智能研究中心
以无监督的方式使用分布密度对数据复杂性进行排序
通过利用课程学习来制定学习策略以有效处理大量噪声标签和数据不均衡问题
利用高噪声标签来提高标准深度网络的模型泛化和整体能力
1
2
3
4
5

6、GML

• Global Meets Local: Effective Multi-Label Image Classification via Category-Aware Weak Supervision

全局满足局部：基于类别感知弱监督的有效多标签图像分类
腾讯优图实验室
1
2

CVPR2022

• A Dual Weighting Label Assignment Scheme for Object Detection

一种用于目标检测的双加权标签分配方案
香港理工大学
双加权的自适应标签分配方案，打破耦合加权的惯例，从不同方面估计一致性和不一致性度量，为每个锚点动态分配单独的正负权重，训练精确的密集目标检测器
新的方框细化操作——直接细化回归图上的方框
但是，会减少训练样本的数量，影响对 small objects 的训练效果
1
2
3
4
5

• ACPL: Anti-curriculum Pseudo-labelling for Semi-supervised Medical Image Classification

ACPL：用于半监督医学图像分类的反课程伪标记
澳大利亚机器学习研究所，阿德莱德大学，德国乌尔姆第二大学
反课程伪标签的 SSL 方法
新的伪标记方法 CSDI —— 不需要估计类分类阈值
IM 集合 KNN 分类器，生成更准确的伪标签
ASP 方法保持密度评分在后续训练阶段的准确性
https://github.com/FBLADL/ACPL
1
2
3
4
5
6
7

• ADeLA_ Automatic Dense Labeling With Attention for Viewpoint Shift in Semantic Segmentation

语义分割中注意力视点转移的自动密集标记
斯坦福、谷歌
将视图转换网络视为（彩色图像隐含的未知映射的）函数表示（a functional representation of an unknown mapping implied by the color images）
提出函数标签幻觉法（functional label hallucination）（在目标域中生成具有不确定性的伪标签）
1
2
3
4

• Back to Reality_ Weakly-Supervised 3D Object Detection With Shape-Guided Label Enhancement

回归现实：基于形状导向标签增强的弱监督 3D 物体检测
清华大学、谷歌
新的标签增强方法 —— BR（Back to Reality），用于只使用对象中心和类标签作为监督训练的 3D 对象检测
https://github.com/wyf-ACCEPT/BackToReality
1
2
3
4

• BoostMIS_ Boosting Medical Image Semi-Supervised Learning With Adaptive Pseudo Labeling and Informative Active Annotation

MoostMIS：使用自适应伪标记和信息性主动性标注促进医学图像半监督学习
新加坡国立大学医学研究院
新的医学图像半监督框架 —— BoostMIS，利用主动学习释放未标记数据的潜力
提出自适应伪标记和信息主动标记，形成闭环
自适应伪标记 —— 动态调整阈值
https://github.com/wannature/BoostMIS
1
2
3
4
5
6

• Cross-Model Pseudo-Labeling for Semi-Supervised Action Recognition

跨模型伪标记的半监督动作识别
香港中文大学，南洋理工大学，微软亚洲研究院
新的半监督动作识别框架 —— 跨模型伪标记框架 CMPL
将轻量级网络与主干网络配对，通过伪标记相互学习
1
2
3
4

• DASO_ Distribution-Aware Semantics-Oriented Pseudo-Label for Imbalanced Semi-Supervised Learning

面向分布感知语义导向的非平衡半监督学习伪标签
韩国科学技术院
通用的伪标记框架，将 “基于相似性的分类器中的语义伪标签” 与 “线性分类器中的线性伪标签” 进行类自适应混合
https://github.com/ytaek-oh/daso
1
2
3
4

• Debiased Learning From Naturally Imbalanced Pseudo-Labels

自然不平衡伪标签的去偏学习
加州大学伯克利分校，微软研究院
通用附加组件 —— 去偏伪标记方案，在不利用任何真实数据分布的先验知识的情况下，动态地减轻偏误伪标签对学生模型的影响
可以集成到半监督学习 SSL 和零镜头学习 ZSL 中的模块
不平衡、长尾分布
https://github.com/frank-xwang/debiased-pseudo-labeling
1
2
3
4
5
6

• Deep Anomaly Discovery From Unlabeled Videos via Normality Advantage and Self-Paced Refinement

1

• Dist-PU_ Positive-Unlabeled Learning From a Label Distribution Perspective

1

• Exploiting Pseudo Labels in a Self-Supervised Learning Framework for Improved Monocular Depth Estimation

1

• FedCorr_ Multi-Stage Federated Learning for Label Noise Correction

1

• Few-Shot Incremental Learning for Label-to-Image Translation

1

• Few-Shot Learning With Noisy Labels

带噪声标签的少样本学习
Facebook AI 研究院，麦吉尔大学
transformer
1
2
3

• Improving Segmentation of the Inferior Alveolar Nerve Through Deep Label Propagation

基于深度标签传播的牙槽神经分割改进
标签传播技术：稀疏注释扩展为密集标签
1
2

• Incorporating Semi-Supervised and Positive-Unlabeled Learning for Boosting Full Reference Image Quality Assessment

结合半监督和正无标记学习促进全参考图像质量评估
https://github.com/happycaoyue/JSPL
1
2

• Incremental Learning in Semantic Segmentation From Image Labels

增量学习在图像标签语义分割中的应用
https://github.com/fcdl94/WILSON
1
2

• Instance-Dependent Label-Noise Learning With Manifold-Regularized Transition Matrix Estimation

基于流形正则化过渡矩阵估计的实例相关标签噪声学习
标签噪声学习
过渡矩阵
统计一致性分类器
1
2
3
4

• Interactive Multi-Class Tiny-Object Detection

交互式多类小目标检测
https://github.com/ChungYi347/Interactive-Multi-Class-Tiny-Object-Detection
1
2

• Label Matching Semi-Supervised Object Detection

标签匹配半监督对象检测
伪标签，均值教师、自适应标签分布
https://github.com/HIK-LAB/SSOD
1
2
3

• Label Relation Graphs Enhanced Hierarchical Residual Network for Hierarchical Multi-Granularity Classification

面向分层多粒度分类的标签关系图增强分层残差网络
https://github.com/MonsterZhZh/HRN
1
2

• Label, Verify, Correct_ A Simple Few Shot Object Detection Method

标记、验证、校正:一种简单的少样本物体检测方法
高质量伪标签
COCO
VOC
物体检测、半监督、少样本
1
2
3
4
5

• Label-Only Model Inversion Attacks via Boundary Repulsion

基于边界排斥的标签模型反转攻击
https://github.com/m-kahla/Label-Only-Model-Inversion-Attacks-via-Boundary-Repulsion
1
2

• Large Loss Matters in Weakly Supervised Multi-Label Classification

弱监督多标签分类中的大损失问题
partial label setting
learn a multi-label classification using partially observed labels per image
the model first learns the representation of clean labels, and then starts memorizing noisy labels
https://github.com/snucml/LargeLossMatters
1
2
3
4
5

• Large-Scale Pre-training for Person Re-identification with Noisy Labels

带噪声标签的人再识别的大规模预训练
https://github.com/DengpanFu/LUPerson-NL
1
2

• Learning Fair Classifiers With Partially Annotated Group Labels

学习带有部分注释的组标签的公平分类器
https://github.com/naver-ai/cgl_fairness
1
2

• Learning From Pixel-Level Noisy Label_ A New Perspective for Light Field Saliency Detection

从像素级噪声标签学习:光场显著性检测的新视角
https://github.com/OLobbCode/NoiseLF
1
2

• Learning To Detect Mobile Objects From LiDAR Scans Without Labels

学习从无标签的激光雷达扫描中检测移动物体
https://github.com/YurongYou/MODEST
1
2

• Learning To Imagine_ Diversify Memory for Incremental Learning Using Unlabeled Data

学习想象:使用无标签数据进行增量学习的多样化记忆
语义对比学习
1
2

• Learning With Neighbor Consistency for Noisy Labels

噪声标签的邻居一致性学习
领域一致性正则化：具有标签噪声的深度学习策略
半监督、转导标签传播
1
2
3

• Learning With Twin Noisy Labels for Visible-Infrared Person Re-Identification

双噪声标签学习的可见-红外人员再识别
https://github.com/XLearning-SCU/2022-CVPR-DART
1
2

• Multi-class Token Transformer for Weakly Supervised Semantic Segmentation

1

• Multidimensional Belief Quantification for Label-Efficient Meta-Learning

1

• Multi-Label Classification With Partial Annotations Using Class-Aware Selective Loss

1

• Multi-Label Iterated Learning for Image Classification With Label Ambiguity

1

• Multi-Marginal Contrastive Learning for Multi-Label Subcellular Protein Localization

1

• Mutual Quantization for Cross-Modal Search With Noisy Labels

1

• Not All Labels Are Equal_ Rationalizing the Labeling Costs for Training Object Detection

1

• Not All Relations Are Equal_ Mining Informative Labels for Scene Graph Generation

1

• On Learning Contrastive Representations for Learning With Noisy Labels

1

• Open-Vocabulary Instance Segmentation via Robust Cross-Modal Pseudo-Labeling

1

• Part-Based Pseudo Label Refinement for Unsupervised Person Re-Identification

1

• PLAD_ Learning To Infer Shape Programs With Pseudo-Labels and Approximate Distributions

1

• PNP_ Robust Learning From Noisy Labels by Probabilistic Noise Prediction

1

• Propagation Regularizer for Semi-Supervised Learning With Extremely Scarce Labeled Samples

1

• Replacing Labeled Real-Image Datasets With Auto-Generated Contours

1

• Safe-Student for Safe Deep Semi-Supervised Learning With Unseen-Class Unlabeled Data

1

• Scalable Penalized Regression for Noise Detection in Learning With Noisy Labels

1

• Selective-Supervised Contrastive Learning With Noisy Labels

1

• Self-Supervised Global-Local Structure Modeling for Point Cloud Domain Adaptation With Reliable Voted Pseudo Labels

1

• Self-Taught Metric Learning without Labels

1

• Semi-Supervised Learning of Semantic Correspondence With Pseudo-Labels

1

• Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labels

1

• The Devil Is in the Labels_ Noisy Label Correction for Robust Scene Graph Generation

1

• The Devil Is in the Margin_ Margin-Based Label Smoothing for Network Calibration

1

• The Neurally-Guided Shape Parser_ Grammar-Based Labeling of 3D Shape Regions With Approximate Inference

1

• Towards Data-Free Model Stealing in a Hard Label Setting

1

• TWIST_ Two-Way Inter-Label Self-Training for Semi-Supervised 3D Instance Segmentation

1

• Undoing the Damage of Label Shift for Cross-Domain Semantic Segmentation

1

• UniCon_ Combating Label Noise Through Uniform Selection and Contrastive Learning

1

• Unified Contrastive Learning in Image-Text-Label Space

1

• Unimodal-Concentrated Loss_ Fully Adaptive Label Distribution Learning for Ordinal Regression

1

• Use All The Labels_ A Hierarchical Multi-Label Contrastive Learning Framework

1

• Which Images To Label for Few-Shot Medical Landmark Detection_

1

other, 2022

• Learning Disentangled Label Representations for Multi-label Classification

学习多标签分类的解纠缠标签表示

1
2

• Global Meets Local: Effective Multi-Label Image Classification via Category-Aware Weak Supervision

全局满足局部：基于类别感知弱监督的有效多标签图像分类
腾讯优图实验室
1
2