Latent Label Research Articles

Deep dual incomplete multi-view multi-label classification via label semantic-guided contrastive learning

Multi-view multi-label learning (MVML) aims to train a model that can explore the multi-view information of the input sample to obtain its accurate predictions of multiple labels. Unfortunately, a majority of existing MVML methods are based on the assumption of data completeness, making them useless in practical applications with partially missing views or some uncertain labels. Recently, many approaches have been proposed for incomplete data, but few of them can handle the case of both missing views and labels. Moreover, these few existing works commonly ignore potentially valuable information about unknown labels or do not sufficiently explore latent label information. Therefore, in this paper, we propose a label semantic-guided contrastive learning method named LSGC for the dual incomplete multi-view multi-label classification problem. Concretely, LSGC employs deep neural networks to extract high-level features of samples. Inspired by the observation of exploiting label correlations to improve the feature discriminability, we introduce a graph convolutional network to effectively capture label semantics. Furthermore, we introduce a new sample-label contrastive loss to explore the label semantic information and enhance the feature representation learning. For missing labels, we adopt a pseudo-label filling strategy and develop a weighting mechanism to explore the confidently recovered label information. We validate the framework on five standard datasets and the experimental results show that our method achieves superior performance in comparison with the state-of-the-art methods.

Beyond the Label Itself: Latent Labels Enhance Semi-supervised Point Cloud Panoptic Segmentation

As the exorbitant expense of labeling autopilot datasets and the growing trend of utilizing unlabeled data, semi-supervised segmentation on point clouds becomes increasingly imperative. Intuitively, finding out more ``unspoken words'' (i.e., latent instance information) beyond the label itself should be helpful to improve performance. In this paper, we discover two types of latent labels behind the displayed label embedded in LiDAR and image data. First, in the LiDAR Branch, we propose a novel augmentation, Cylinder-Mix, which is able to augment more yet reliable samples for training. Second, in the Image Branch, we propose the Instance Position-scale Learning (IPSL) Module to learn and fuse the information of instance position and scale, which is from a 2D pre-trained detector and a type of latent label obtained from 3D to 2D projection. Finally, the two latent labels are embedded into the multi-modal panoptic segmentation network. The ablation of the IPSL module demonstrates its robust adaptability, and the experiments evaluated on SemanticKITTI and nuScenes demonstrate that our model outperforms the state-of-the-art method, LaserMix.

LSFSR: Local label correlation-based sparse multilabel feature selection with feature redundancy

In recent studies, existing multilabel feature selection models have focused on either considering the relationship between labels or the redundancy between features. Furthermore, they only use simple sparsity constraints to process high-dimensional data without the intrinsic relationships between features and labels. These issues can have a great impact on the classification effectiveness of feature selection. To address these limitations, this article describes a new local label correlation-based sparse multilabel feature selection approach with feature redundancy. First, a new loss function is established among the matrices of samples, label coefficients, and labels. Then, the Frobenius norm is imposed to investigate the potential relationships between features and labels. The weight matrix is sparsified by the l2,1 norm to ensure that the new loss function has high interpretability. Second, a manifold constraint is employed to capture the local geometric structure between labels and to delve deeper into the latent information among the local labels. Then manifold constraints and Laplacian scores are combined for embedding feature selection to guide the exploration of hidden latent label. Finally, by considering the differences between the feature scores and the redundancy between the samples, feature redundancy is analyzed via the modified cosine similarity, and a candidate feature subset with low redundancy is generated. The l2 norm is used to select features with low redundancy while preserving sparsity, and a novel objective function is developed to optimize this solution. Thus, a sparse feature selection algorithm via local label correlation and feature redundancy is designed, and has demonstrated remarkable classification effectiveness in comparative experiments conducted on 21 multilabel datasets.

Multilabel Feature Selection via Shared Latent Sublabel Structure and Simultaneous Orthogonal Basis Clustering.

Multilabel feature selection solves the dimension distress of high-dimensional multilabel data by selecting the optimal subset of features. Noisy and incomplete labels of raw multilabel data hinder the acquisition of label-guided information. In existing approaches, mapping the label space to a low-dimensional latent space by semantic decomposition to mitigate label noise is considered an effective strategy. However, the decomposed latent label space contains redundant label information, which misleads the capture of potential label relevance. To eliminate the effect of redundant information on the extraction of latent label correlations, a novel method named SLOFS via shared latent sublabel structure and simultaneous orthogonal basis clustering for multilabel feature selection is proposed. First, a latent orthogonal base structure shared (LOBSS) term is engineered to guide the construction of a redundancy-free latent sublabel space via the separated latent clustering center structure. The LOBSS term simultaneously retains latent sublabel information and latent clustering center structure. Moreover, the structure and relevance information of nonredundant latent sublabels are fully explored. The introduction of graph regularization ensures structural consistency in the data space and latent sublabels, thus helping the feature selection process. SLOFS employs a dynamic sublabel graph to obtain a high-quality sublabel space and uses regularization to constrain label correlations on dynamic sublabel projections. Finally, an effective convergence provable optimization scheme is proposed to solve the SLOFS method. The experimental studies on the 18 datasets demonstrate that the presented method performs consistently better than previous feature selection methods.

Discriminative Regression With Latent Label Learning for Image Classification

Discriminative Regression With Latent Label Learning for Image Classification

Transformed Schatten-1 penalty based full-rank latent label learning for incomplete multi-label classification

Transformed Schatten-1 penalty based full-rank latent label learning for incomplete multi-label classification

Missing multi-label learning based on the fusion of two-level nonlinear mappings

The relationship between features and labels is an important factor affecting multi-label learning. Many existing multi-label learning approaches decompose label space into a low-rank latent label space and construct linear mappings between the feature space and the latent label space. This type of models may not depict the true data distribution well in high-dimensional spaces. In this work, a model called missing multi-label learning based on the fusion of two-level nonlinear mappings (MML-TLNM) is proposed. Specifically, in the second-level mapping, the label matrix is nonlinearly decomposed into a low-rank latent label matrix and a correlation coefficient matrix to realize the nonlinear transformation from the latent label space to the actual label space. In the first-level mapping, a nonlinear relationship is constructed from the sample matrix to the latent label matrix to realize the nonlinear transformation between the original feature space and the latent label space. The two levels of nonlinear mappings are fused with each other in a nested manner and better depict the true distribution of multi-label data. In addition, the F-norm regularization constraints are imposed on the variables in the two-level nonlinear mappings. Extensive comparative experiments showed that the proposed learning model has good robustness and is superior to the state-of-the-art multi-label learning methods.

Multi-View Partial Multi-Label Learning via Graph-Fusion-Based Label Enhancement

Multi-view partial multi-label learning (MVPML) aims to learn a multi-label predictive model from the training examples, each of which is presented by multiple feature vectors while associated with a set of candidate labels where only a subset is correct. Generally, existing techniques work simply by identifying the ground-truth label via aggregating the features from all views to train a final classifier, but ignore the cause of the incorrect labels in the candidate label sets, i.e., the diverse property of the representation from different views leads to the incorrect labels which form the candidate labels alone with the essential supervision. In this paper, a novel MVPML approach is proposed to learn the predictive model and the incorrect-labeling model jointly by incorporating the graph-fusion-based topological structure of the feature space. Specifically, the latent label distribution and the incorrect labels are identified simultaneously in a unified framework under the supervision of candidate labels. In addition, a common topological structure of the feature space from all views is learned via the graph fusion for further capturing the latent label distribution. Experimental results on the real-world datasets clearly validate the effectiveness of the proposed approach for solving multi-view partial multi-label learning problems.

Learning Common and Task-Specific Radiomic Features via Graph Regularized NMF for the Joint Prediction of Multiple Clinical Indicators in Breast Cancer.

Assessments of multiple clinical indicators based on radiomic analysis of magnetic resonance imaging (MRI) are beneficial to the diagnosis, prognosis and treatment of breast cancer patients. Many machine learning methods have been designed to jointly predict multiple indicators for more accurate assessments while using original clinical labels directly without considering the noisy and redundant information among them. To this end, we propose a multilabel learning method based on label space dimensionality reduction (LSDR), which learns common and task-specific features via graph regularized nonnegative matrix factorization (CTFGNMF) for the joint prediction of multiple indicators in breast cancer. A nonnegative matrix factorization (NMF) is adopted to map original clinical labels to a low-dimensional latent space. The latent labels are employed to exploit task correlations by using a least square loss function with [Formula: see text]-norm regularization to identify common features, which help to improve the generalization performance of correlated tasks. Furthermore, task-specific features were retained by a multitask regression formulation to increase the discrimination power for different tasks. Common and task-specific features are incorporated by dynamic graph Laplacian regularization into a unified model to learn complementary features. Then, a multilabel classification is built to predict multiple clinical indicators including human epidermal growth factor receptor 2 (HER2), Ki-67, and histological grade. Experimental results show that CTFGNMF achieves AUCs of 0.823, 0.691 and 0.776 in the three indicator predictions, outperforming other counterparts that consider only task-independent features or common features. It indicates CTFGNMF is a promising application for multiple classification tasks in breast cancer.

A nonlinear multi-label learning model based on Tanh mapping

The relationship between features and labels plays an important role in multi-label learning. The purpose of multi-label learning is to learn a mapping from the feature space to the label space. Many of the existing methods decompose the observed label matrix into the product of a low-rank latent label matrix and a projection matrix, and construct a linear mapping between the feature space and low-rank latent label space. However, in practice, the relationship between two high-dimensional spaces is often nonlinear. Consequently, these linear models may not sufficiently capture the data distribution and the true dependencies between features and labels. Herein, a novel classification model for multi-label learning is proposed, which considers nonlinear mapping from the feature space to the latent label space. Because of correlation between labels, the label matrix is usually of a low rank. To explore the low-rank structure, a latent label matrix is obtained using the decomposition of the original label matrix. To characterize the nonlinear relationship between features and latent labels, the tanh function is employed to learn a nonlinear classifier. The tanh function describes the distribution of labels by constraining the domain of the latent label space to the interval [−1,1]. In addition, the Frobenius norm regularization constraint is imposed on the variables in the model for optimization. The experimental results demonstrate that the proposed method is robust and comparable to state-of-the-art multi-label learning methods.

Progressive Enhancement of Label Distributions for Partial Multilabel Learning.

Partial multi-label learning (PML) aims to learn a multilabel predictive model from the PML training examples, each of which is associated with a set of candidate labels where only a subset is valid. The common strategy to induce a predictive model is identifying the valid labels in each candidate label set. Nonetheless, this strategy ignores considering the essential label distribution corresponding to each instance as label distributions are not explicitly available in the training dataset. In this article, a novel partial multilabel learning method is proposed to recover the latent label distribution and progressively enhance it for predictive model induction. Specifically, the label distribution is recovered by considering the observation model for logical labels and the sharing topological structure from feature space to label distribution space. Besides, the latent label distribution is progressively enhanced by recovering latent labeling information and supervising predictive model training alternatively to make the label distribution appropriate for the induced predictive model. Experimental results on PML datasets clearly validate the effectiveness of the proposed method for solving partial multilabel learning problems. In addition, further experiments show the high quality of the recovered label distributions and the effectiveness of adopting label distributions for partial multilabel learning.

GCANet: Geometry cues-aware facial expression recognition based on graph convolutional networks

Facial expression recognition (FER) task in the wild is challenging due to some uncertainties, such as the ambiguity of facial expressions, subjective annotations, and low-quality facial images. A novel model for FER in-the-wild datasets is proposed in this study to solve these uncertainties. The overview of the proposed method is as follows. First, the facial images are grouped into high and low uncertainties by the pre-trained network. The graph convolutional network (GCN) framework is then used for the facial images with low uncertainty to obtain geometry cues, including the relationship among action units (AUs) and the implicit connection between AUs and expressions, which help predict the probability of the underlying emotional label. The emotion label distribution is produced by combining the predicted latent label probability and the given label. For the facial images with high uncertainty, k-nearest neighbor graphs are built to determine the k facial images in the low uncertainty group with the highest similarity to the given facial image. The emotion label distribution of the given image is then replaced by fusing the emotion label distribution based on the distances between the given image and its adjacent images. Finally, the constructed emotion label distribution facilitates training in a straightforward manner using a convolutional neural network framework to identify facial expressions. Experimental results on RAF-DB, FERPlus, AffectNet, and SFEW2.0 datasets demonstrate that the proposed method achieved superior performance compared to state-of-the-art approaches.

Multilabel Feature Selection With Constrained Latent Structure Shared Term.

High-dimensional multilabel data have increasingly emerged in many application areas, suffering from two noteworthy issues: instances with high-dimensional features and large-scale labels. Multilabel feature selection methods are widely studied to address the issues. Previous multilabel feature selection methods focus on exploring label correlations to guide the feature selection process, ignoring the impact of latent feature structure on label correlations. In addition, one encouraging property regarding correlations between features and labels is that similar features intend to share similar labels. To this end, a latent structure shared (LSS) term is designed, which shares and preserves both latent feature structure and latent label structure. Furthermore, we employ the graph regularization technique to guarantee the consistency between original feature space and latent feature structure space. Finally, we derive the shared latent feature and label structure feature selection (SSFS) method based on the constrained LSS term, and then, an effective optimization scheme with provable convergence is proposed to solve the SSFS method. Better experimental results on benchmark datasets are achieved in terms of multiple evaluation criteria.

Generative Multi-Label Correlation Learning

In real-world applications, a single instance could have more than one label. To solve this task, multi-label learning methods emerged in recent years. It is a more challenging problem for many reasons, such as complex label correlation, long-tail label distribution, and data shortage. In general, overcoming these challenges and bettering learning performance could be achieved by utilizing more training samples and including label correlations. However, these solutions are expensive and inflexible. Large-scale, well-labeled datasets are difficult to obtain, and building label correlation maps requires task-specific semantic information as prior knowledge. To address these limitations, we propose a general and compact Multi-Label Correlation Learning (MUCO) framework. MUCO explicitly and effectively learns the latent label correlations by updating a label correlation tensor, which provides highly accurate and interpretable prediction results. In addition, a multi-label generative strategy is deployed to handle the long-tail label distribution challenge. It borrows the visual clues from limited samples and synthesizes more diverse samples. All networks in our model are optimized simultaneously. Extensive experiments illustrate the effectiveness and efficiency of MUCO. Ablation studies further prove the effectiveness of all the modules.

Uncertainty-Aware Scene Graph Generation

As a challenging task in computer vision, Scene graph generation (SGG) aims to model the underlying semantic relationships among objects in a given image for scene understanding. Due to the increasing scale and subjectivity, the annotations of existing SGG benchmarks inevitably suffer from some uncertainty issues, resulting in the models hardly learning the relationships comprehensively. In this work, we address the uncertainty from the perspectives of both classifier parameters and relationship labels. On one hand, we handle the classifier uncertainty via learning a Bayesian classifier reparameterization, of which the weights are sampled from a latent space spanned with a prior distribution. On the other hand, we assume that each relationship label is sampled from a latent label space and mitigate the label uncertainty via estimating the latent relationship distribution. As a result, the distribution of the classifier parameters are comprehensively learned under the supervision of the estimated relationship labels, thus improving the model’s generalization ability. Experimental results on the popular benchmark demonstrate that the proposed strategies significantly improve different baseline models on different SGG tasks.

Nonlinear industrial process fault diagnosis with latent label consistency and sparse Gaussian feature learning

Nonlinear industrial process fault diagnosis with latent label consistency and sparse Gaussian feature learning

Label prediction based constrained non-negative matrix factorization for semi-supervised multi-view classification

Semi-supervised multi-view classification can improve the performance by leveraging the information from both labeled and unlabeled data. But it is often a challenge to capture the information from the unlabeled multi-view data. By analyzing the relation between labeled and unlabeled data under multi-view scenario, we propose a novel model with the ability of leveraging the latent label information from the unlabeled data. In our model, a label prediction (LP) term is proposed to jointly obtain the predicted labels of unlabeled data from multiple views. The LP term is integrated into a constrained non-negative matrix factorization based multi-view framework. In this way, the LP and the multi-view representation learning are integrated into one joint learning problem, where they boost each other. Particularly, the predicted label vector is formulated to be the one-hot vector, such that the labels can be obtained directly. Moreover, we propose a new lemma about the gradient of the ℓ2,1 norm in the case of 3-factor matrix decomposition and its corollary about multi-factor matrix decomposition. Based on which, we develop an efficient algorithm and prove its convergence. Experimental results verify that our method can obtain state-of-the-art performance.

Robust Multi-Label Classification with Enhanced Global and Local Label Correlation

Data representation is of significant importance in minimizing multi-label ambiguity. While most researchers intensively investigate label correlation, the research on enhancing model robustness is preliminary. Low-quality data is one of the main reasons that model robustness degrades. Aiming at the cases with noisy features and missing labels, we develop a novel method called robust global and local label correlation (RGLC). In this model, subspace learning reconstructs intrinsic latent features immune from feature noise. The manifold learning ensures that outputs obtained by matrix factorization are similar in the low-rank latent label if the latent features are similar. We examine the co-occurrence of global and local label correlation with the constructed latent features and the latent labels. Extensive experiments demonstrate that the classification performance with integrated information is statistically superior over a collection of state-of-the-art approaches across numerous domains. Additionally, the proposed model shows promising performance on multi-label when noisy features and missing labels occur, demonstrating the robustness of multi-label classification.

Multi-label learning with missing features and labels and its application to text categorization

• This paper proposes a new method MMFL to address the problem of missing features and missing labels simultaneously in multi-label learning. • Matrix decomposition technology is used to recover the missing values of features and labels. In addition, to overcome the problem of tail labels in matrix factorization, an extra classifier for the sparse tail labels is built. • The experiment on nine multi-label text categorization data sets show that our proposed method achieves a competitive performance in multi-label learning with missing features and labels. In multi-label learning, researchers usually assume that the training data set is complete. However, this assumption is not always hold in real applications, e.g., the features or labels are missing for some data examples. Existing algorithms mainly focus on the problem of missing labels, and ignore the problem of missing features. This paper proposes a novel multi-label learning algorithm named MMFL, i.e., Multi-label learning with Missing Features and Labels, which can deal with the problem of missing features and labels simultaneously. First, we try to recover the missing values of features and labels by matrix factorization, and then learn a classification model from the latent feature space to the latent label space. Second, to overcome the problem of tail labels in matrix factorization, we build an extra classifier for the sparse tail labels. Besides, the manifold regularization technology is used to keep the manifold structures of instance similarity and label correlation. The effectiveness of our proposed method is verified by comparing it with the state-of-the-art approaches over eight multi-label benchmark data sets.

Probabilistic online self-distillation

Deploying state-of-the-art deep learning models on devices with limited computational power imposes certain computation and storage restrictions. Knowledge Distillation, i.e. training compact models by transferring knowledge from more powerful models, constitutes a promising route to address this issue that has been followed during the recent years. A limitation of conventional knowledge distillation is that it is a long-lasting, computationally and memory demanding process, since it requires multiple stages of training process. To this end, a novel online probabilistic self-distillation method, namely Probabilistic Online Self-Distillation (POSD), aiming to improve the performance of any deep neural model in an online manner, is proposed in this paper. We argue that considering a classification problem, apart from the explicit concepts expressed with the hard labels, there are also implicit concepts expressed with the so-called latent labels. These implicit concepts reflect similarities among data, regardless of the classes. Then, our goal is to maximize the Mutual Information between the data samples and the latent labels. In this way, we are able to derive additional knowledge from the model itself, without the need of building multiple identical models or using multiple models to teach each other, like existing online distillation methods, rendering the POSD method more efficient. The experimental evaluation on six datasets validates that the proposed method improves the classification performance.