Learning from Ambiguously Labeled Face Images

Chen, Ching‐Hui; Patel, Vishal M.; Chellappa, Rama

doi:10.1109/tpami.2017.2723401

Cited by 89 publications

(32 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In partial label (PL) learning, each training example is represented by a single instance (feature vector) while associated with a set of candidate labels, only one of which is the ground-truth label. This learning paradigm is also termed as superset label learning (Liu and Dietterich 2012;Liu and Dietterich 2014;Hüllermeier and Cheng 2015;Gong et al 2018) or ambiguous label learning (Hüllermeier and Beringer 2006;Zeng et al 2013;Chen et al 2014;Chen, Patel, and Chellappa 2017). Since manually labeling the ground-truth label of each instance could incur unaffordable monetary or time cost, partial label learning has various application domains, such as web mining (Luo and Orabona 2010), image annotation (Cour, Sapp, and Taskar 2011;Zeng et al 2013), and ecoinformatics (Liu and Dietterich 2012).…”

Section: Introductionmentioning

confidence: 99%

Partial Label Learning with Self-Guided Retraining

Feng

2019

AAAI

View full text Add to dashboard Cite

Partial label learning deals with the problem where each training instance is assigned a set of candidate labels, only one of which is correct. This paper provides the first attempt to leverage the idea of self-training for dealing with partially labeled examples. Specifically, we propose a unified formulation with proper constraints to train the desired model and perform pseudo-labeling jointly. For pseudo-labeling, unlike traditional self-training that manually differentiates the ground-truth label with enough high confidence, we introduce the maximum infinity norm regularization on the modeling outputs to automatically achieve this consideratum, which results in a convex-concave optimization problem. We show that optimizing this convex-concave problem is equivalent to solving a set of quadratic programming (QP) problems. By proposing an upper-bound surrogate objective function, we turn to solving only one QP problem for improving the optimization efficiency. Extensive experiments on synthesized and real-world datasets demonstrate that the proposed approach significantly outperforms the state-of-the-art partial label learning approaches.

show abstract

Section: Introductionmentioning

confidence: 99%

Partial Label Learning with Self-Guided Retraining

Feng

2019

AAAI

View full text Add to dashboard Cite

show abstract

“…In other words, the specific correspondences between the faces and their names are unknown. In addition to the common scenarios mentioned above, PLL has also achieved competitive performance in many other applications, such as multimedia content analysis [7] [8] [9] [10], facial age estimation [11], web mining [12], ecoinformatics [13], etc.…”

Section: Introductionmentioning

confidence: 99%

GM-PLL: Graph Matching Based Partial Label Learning

Lyu

Feng

Wang

et al. 2021

IEEE Trans. Knowl. Data Eng.

View full text Add to dashboard Cite

Partial Label Learning (PLL) aims to learn from the data where each training example is associated with a set of candidate labels, among which only one is correct. The key to deal with such problem is to disambiguate the candidate label sets and obtain the correct assignments between instances and their candidate labels. In this paper, we interpret such assignments as instance-to-label matchings, and reformulate the task of PLL as a matching selection problem. To model such problem, we propose a novel Graph Matching based Partial Label Learning (GM-PLL) framework, where Graph Matching (GM) scheme is incorporated owing to its excellent capability of exploiting the instance and label relationship. Meanwhile, since conventional one-to-one GM algorithm does not satisfy the constraint of PLL problem that multiple instances may correspond to the same label, we extend a traditional one-to-one probabilistic matching algorithm to the many-to-one constraint, and make the proposed framework accommodate to the PLL problem. Moreover, we also propose a relaxed matching prediction model, which can improve the prediction accuracy via GM strategy. Extensive experiments on both artificial and real-world data sets demonstrate that the proposed method can achieve superior or comparable performance against the state-of-the-art methods.

show abstract

“…How to effectively learn with such data has attracted much attention in the data mining community. Formally, the paradigm is referred to as Partial Label (PL) learning [6-9, 15, 27, 28, 37], also superset label learning [11,16,17] or ambiguous label learning [2,3,5,13,33]. The PL scenarios are mainly caused by the expensive cost of acquiring explicit labels for instances, and they emerge in many real-world applications, e.g., multimedia content analysis [5,33], web mining [18], and ecoinformatics [16], etc.…”

Section: Introductionmentioning

confidence: 99%

Learning with Noisy Partial Labels by Simultaneously Leveraging Global and Local Consistencies

Ouyang

2020

Proceedings of the 29th ACM International Conference on Information &Amp; Knowledge Management

View full text Add to dashboard Cite

In real-world scenarios, the data are widespread that are annotated with a set of candidate labels but a single ground-truth label perinstance. The learning paradigm with such data, formally referred to as Partial Label (PL) learning, has recently drawn much attention. The traditional PL methods estimate the confidences being the ground-truth label of candidate labels with various regularizations and constraints, however, they only consider the local information, resulting in potentially less accurate estimations as well as worse classification performance. To alleviate this problem, we propose a novel PL method, namely PArtial label learN ing by simultaneously leveraging GlObal and Local consI steN cies (Pangolin). Specifically, we design a global consistency regularization term to pull instances associated with similar labeling confidences together by minimizing the distances between instances and label prototypes, and a local consistency term to push instances marked with no same candidate labels away by maximizing their distances. We further propose a nonlinear kernel extension of Pangolin, and employ the Taylor approximation trick for efficient optimization. Empirical results demonstrate that Pangolin significantly outperforms the existing PL baseline methods. CCS CONCEPTS • Computing methodologies → Machine learning algorithms; • Information systems → Clustering and classification.

show abstract

Learning from Ambiguously Labeled Face Images

Cited by 89 publications

References 28 publications

Partial Label Learning with Self-Guided Retraining

Partial Label Learning with Self-Guided Retraining

GM-PLL: Graph Matching Based Partial Label Learning

Learning with Noisy Partial Labels by Simultaneously Leveraging Global and Local Consistencies

Contact Info

Product

Resources

About