Common Representation Learning Research Articles

Heterogeneous Federated Domain Generalization Network With Common Representation Learning for Cross-Load Machinery Fault Diagnosis

Heterogeneous Federated Domain Generalization Network With Common Representation Learning for Cross-Load Machinery Fault Diagnosis

M2R2: Missing-Modality Robust Emotion Recognition Framework With Iterative Data Augmentation

This paper deals with the utterance-level modalities missing problem with uncertain patterns on emotion recognition in conversation (ERC) task. Present models generally predict the speaker's emotions by its current utterance and context, which is degraded by modality missing considerably. Our work proposes a framework Missing-Modality Robust emotion Recognition (M2R2), which trains emotion recognition model with iterative data augmentation by learned common representation. Firstly, a network called Party Attentive Network (PANet) is designed to classify emotions, which tracks all the speakers' states and context. Attention mechanism between speaker with other participants and dialogue topic is used to decentralize dependence on multi-time and multi-party utterances instead of the possible incomplete one. Moreover, the Common Representation Learning (CRL) problem is defined for modality-missing problem. Data imputation methods improved by the adversarial strategy are used here to construct extra features to augment data. Extensive experiments and case studies validate the effectiveness of our methods over baselines for modality-missing emotion recognition on two different datasets.

Maximizing Mutual Information Across Feature and Topology Views for Representing Graphs

Recently, maximizing mutual information has emerged as a powerful tool for unsupervised graph representation learning. Existing methods are typically effective in capturing graph information from the topology view but consistently ignore the node feature view. To circumvent this problem, we propose a novel method by exploiting mutual information maximization across feature and topology views. Specifically, we first construct the feature graph to capture the underlying structure of nodes in feature spaces by measuring the distance between pairs of nodes. Then we use a cross-view representation learning module to capture both local and global information content across feature and topology views on graphs. To model the information shared by the feature and topology spaces, we develop a common representation learning module by using mutual information maximization and reconstruction loss minimization. Here, minimizing reconstruction loss forces the model to learn the shared information of feature and topology spaces. To explicitly encourage diversity between graph representations from the same view, we also introduce a disagreement regularization to enlarge the distance between representations from the same view. Experiments on synthetic and real-world datasets demonstrate the effectiveness of integrating feature and topology views. In particular, compared with the previous supervised methods, the proposed method achieves comparable or even better performance under the unsupervised representation and linear evaluation protocol.

Common latent representation learning for low-resourced spoken language identification

Common latent representation learning for low-resourced spoken language identification

Multi-view unsupervised feature selection with tensor robust principal component analysis and consensus graph learning

Recently, multi-view unsupervised feature selection has attracted much attention due to its efficiency and better interpretability in processing high-dimensional multi-view datasets. Most existing methods rely on the constructed similarity matrices to obtain reliable pseudo labels to guide the feature selection. However, the considerable adverse noise in the raw data inevitably impedes the exploration of true underlying similarity structures. Besides, the inter-view correlations are often ignored during the common representation learning, which limits the effective fusion of the essential information from multiple views. To solve these issues, we design a novel robust multi-view unsupervised feature selection framework. Specifically, our method seeks a set of noise-free view-specific similarity matrices by leveraging tensor robust principal component analysis, where the high-order connections among different views are well exploited through the constructed low-rank tensor. Meanwhile, a high-quality consensus similarity matrix is adaptively learned from the view-specific representations within the same unified framework to capture the shared local structures. To enhance the discriminative ability of the feature selection matrix, we further impose a rank constraint on the consensus similarity matrix to obtain reliable pseudo cluster indicators. We present an efficient optimization algorithm ground on the alternating direction method of multipliers to solve the proposed model. Experimental results on six multi-view datasets confirm the superiority of our method.

Nearest Neighbor-Based Strategy to Optimize Multi-View Triplet Network for Classification of Small-Sample Medical Imaging Data.

Multi-view classification with limited sample size and data augmentation is a very common machine learning (ML) problem in medicine. With limited data, a triplet network approach for two-stage representation learning has been proposed. However, effective training and verifying the features from the representation network for their suitability in subsequent classifiers are still unsolved problems. Although typical distance-based metrics for the training capture the overall class separability of the features, the performance according to these metrics does not always lead to an optimal classification. Consequently, an exhaustive tuning with all feature-classifier combinations is required to search for the best end result. To overcome this challenge, we developed a novel nearest-neighbor (NN) validation strategy based on the triplet metric. This strategy is supported by a theoretical foundation to provide the best selection of the features with a lower bound of the highest end performance. The proposed strategy is a transparent approach to identify whether to improve the features or the classifier. This avoids the need for repeated tuning. Our evaluations on real-world medical imaging tasks (i.e., radiation therapy delivery error prediction and sarcoma survival prediction) show that our strategy is superior to other common deep representation learning baselines [i.e., autoencoder (AE) and softmax]. The strategy addresses the issue of feature's interpretability which enables more holistic feature creation such that the medical experts can focus on specifying relevant data as opposed to tedious feature engineering.

One-step incomplete multiview clustering with low-rank tensor graph learning

Incomplete multiview clustering (IMVC) has attracted extensive attention in the field of machine learning due to its excellent performance in handling incomplete multiview data (IMVD). However, existing IMVC methods often use spectral clustering or k-means on the obtained affinity matrices to obtain the final clustering results, that is, the affinity matrices learning process and clustering process are separated. In this paper, we propose a new ingenious IMVC method, one-step incomplete multiview clustering with low-rank tensor graph learning (OIMVC/LTGL), to solve the problem. This method combines graph learning, low-rank tensor constraint, common representation learning and clustering into a unified optimization framework. First, we utilize the adjacency relationship between views to complete the similarity graph matrices. Second, we stack all similarity graph matrices into a third-order tensor. To mine higher-order correlations among different views, we impose a low-rank tensor constraint, the tensor nuclear norm (TNN), into the constructed tensor. Third, to explore consistent information between views, we introduce a common representation learning term to learn the optimal consensus representation. Last, we introduce spectral rotation to the consensus representation matrices to directly obtain the clustering labels. These steps promote each other for better clustering performance. We solve the optimization problem by using the augmented Lagrange multiplier (ALM) method. Experiments based on five well-known datasets show the superiority of our method.

Semi-supervised cross-modal retrieval with graph-based semantic alignment network

Semi-supervised cross-modal retrieval is an eclectic paradigm which learns common representations via exploiting underlying semantic information from both labeled and unlabeled data. Most existing methods ignore the rich semantic information of text data and are unable to fully utilize the text data in common representation learning. Moreover, they only considered the correlation of the data with the same semantic label, but ignored the correlation between the data with different semantic label. In this paper, we propose a novel semi-supervised cross-modal retrieval method, called Graph-based Semantic Alignment Network (GSAN), which learns common representation by aligning the features of different modalities with semantic embeddings of text data. Firstly, we design a Deep Supervised Semantic Encoding (DSSE) module to train the semantic projector and label predictor which can exploit the semantic embeddings and the predicted labels from unlabeled data of text modality. Then, GAN-based Bidirectional Fusion (GBF) module is designed to learn the mapping networks of two modalities (image and text). In order to make the mapping networks generate semantically discriminative and modality-invariant representations, we utilize the underlying semantic information exploited by DSSE to construct Graph-based Triplet Constraint (GTC) which can enforce feature embeddings from the semantically-matched (image and text) pairs to be more similar and push those mismatched ones away. By the benefit of fully using of semantic information, our approach can only use fewer label data and achieves the performance of state-of-the-art methods. In addition, since we only utilize the mapping networks trained in GBF module to generate common representations in referring stage, our approach is efficient and time saving in real world application. Extensive experiments on four widely-used datasets show the effectiveness of GSAN.

Self-Supervised Pre-Training with Bridge Neural Network for SAR-Optical Matching

Due to the vast geometric and radiometric differences between SAR and optical images, SAR-optical image matching remains an intractable challenge. Despite the fact that the deep learning-based matching model has achieved great success, SAR feature embedding ability is not fully explored yet because of the lack of well-designed pre-training techniques. In this paper, we propose to employ the self-supervised learning method in the SAR-optical matching framework, in order to serve as a pre-training strategy for improving the representation learning ability of SAR images as well as optical images. We first use a state-of-the-art self-supervised learning method, Momentum Contrast (MoCo), to pre-train an optical feature encoder and an SAR feature encoder separately. Then, the pre-trained encoders are transferred to an advanced common representation learning model, Bridge Neural Network (BNN), to project the SAR and optical images into a more distinguishable common feature representation subspace, which leads to a high multi-modal image matching result. Experimental results on three SAR-optical matching benchmark datasets show that our proposed MoCo pre-training method achieves a high matching accuracy up to 0.873 even for the complex QXS-SAROPT SAR-optical matching dataset. BNN pre-trained with MoCo outperforms BNN with the most commonly used ImageNet pre-training, and achieves at most 4.4% gains in matching accuracy.

Boosting Ship Detection in SAR Images With Complementary Pretraining Techniques

Deep learning methods have made significant progress in ship detection in synthetic aperture radar (SAR) images. The pretraining technique is usually adopted to support deep neural networks-based SAR ship detectors due to the scarce labeled SAR images. However, directly leveraging ImageNet pretraining is hard to obtain a good ship detector because of different imaging perspectives and geometry. In this article, to resolve the problem of inconsistent imaging perspectives between ImageNet and earth observations, we propose an optical ship detector (OSD) pretraining technique to transfer the characteristics of ships in earth observations to SAR images from a large-scale aerial image dataset. On the other hand, to handle the problem of different imaging geometry between optical and SAR images, we propose an optical-SAR matching (OSM) pretraining technique, which transfers plentiful texture features from optical images to SAR images by common representation learning on the OSM task. Finally, observing that the OSD pretraining-based SSD has a better recall on sea area while the OSM pretraining-based SSD can reduce false alarms on land area, we combine the predictions of the two detectors through weighted boxes fusion to further improve detection results. Extensive experiments on four SAR ship detection datasets and three representative convolutional neural network-based detection benchmarks are conducted to show the effectiveness and complementarity of the two proposed detectors, and the state-of-the-art performance of the combination of the two detectors. The proposed method won the sixth place of ship detection in SAR images in the 2020 Gaofen challenge.

Zero-Shot Deep Domain Adaptation With Common Representation Learning.

Domain Adaptation aims at adapting the knowledge learned from a domain (source-domain) to another (target-domain). Existing approaches typically require a portion of task-relevant target-domain data a priori. We propose an approach, zero-shot deep domain adaptation (ZDDA), which uses paired dual-domain task-irrelevant data to eliminate the need for task-relevant target-domain training data. ZDDA learns to generate common representations for source and target domains data. Then, either domain representation is used later to train a system that works on both domains or having the ability to eliminate the need to either domain in sensor fusion settings. Two variants of ZDDA have been developed: ZDDA for classification task (ZDDA-C) and ZDDA for metric learning task (ZDDA-ML). Another limitation in Existing approaches is that most of them are designed for the closed-set classification task, i.e., the sets of classes in both the source and target domains are "known." However, ZDDA-C is also applicable to the open-set classification task where not all classes are "known" during training. Moreover, the effectiveness of ZDDA-ML shows ZDDA's applicability is not limited to classification tasks. ZDDA-C and ZDDA-ML are tested on classification and metric-learning tasks, respectively. Under most experimental conditions, ZDDA outperforms the baseline without using task-relevant target-domain-training data.

Multivariate Extreme Learning Machine Based AutoEncoder for Electricity Consumption Series Clustering

Multivariate electricity consumption series clustering can reflect the trend of power consumption changes in the past time period, which can provide reliable guidance for electricity production. The dimensionality reduction-based method is an effective technology to address this problem, which obtains the low-dimensional features of each variate or all variates for multivariate time series clustering. However, most existing dimensionality reduction-based methods ignore the joint learning of the common representations and the variable-based representations. In this paper, we build a multivariate extreme learning machine based autoencoder model for electricity consumption clustering (MELM-EC), which performs common representation learning and variable-based representation learning simultaneously. MELM-EC maps the common representation and multiple variable-based representations to the original multivariate time series and computes the common output weights within a few iterations. Experimental results on realistic multivariate time series datasets and multivariate electricity consumption series datasets demonstrate the effectiveness of the proposed MELM-EC model.

Semi-supervised cross-modal representation learning with GAN-based Asymmetric Transfer Network

In this paper, we proposed a semi-supervised common representation learning method with GAN-based Asymmetric Transfer Network (GATN) for cross modality retrieval. GATN utilizes the asymmetric pipeline to guarantee the semantic consistency and adopt (Generative Adversarial Network) GAN to fit the distributions of different modalities. Specifically, the common representation learning across modalities includes two stages: (1) the first stage, GATN trains source mapping network to learn the semantic representation of text modality by supervised method; and (2) the second stage, GAN-based unsupervised modality transfer method is proposed to guide the training of target mapping network, which includes generative network (target mapping network) and discriminative network. Experimental results on three widely-used benchmarks show that GATN have achieved better performance comparing with several existing state-of-the-art methods.

Learning Cross-Modal Common Representations by Private-Shared Subspaces Separation.

Due to the inconsistent distributions and representations of different modalities (e.g., images and texts), it is very challenging to correlate such heterogeneous data. A standard solution is to construct one common subspace, where the common representations of different modalities are generated to bridge the heterogeneity gap. Existing methods based on common representation learning mostly adopt a less effective two-stage paradigm: first, generating separate representations for each modality by exploiting the modality-specific properties as the complementary information, and then capturing the cross-modal correlation in the separate representations for common representation learning. Moreover, these methods usually neglect that there may exist interference in the modality-specific properties, that is, the unrelated objects and background regions in images or the noisy words and incorrect sentences in the text. In this article, we hypothesize that explicitly modeling the interference within each modality can improve the quality of common representation learning. To this end, we propose a novel model private-shared subspaces separation (P3S) to explicitly learn different representations that are partitioned into two kinds of subspaces: 1) the common representations that capture the cross-modal correlation in a shared subspace and 2) the private representations that model the interference within each modality in two private subspaces. By employing the orthogonality constraints between the shared subspace and the private subspaces during the one-stage joint learning procedure, our model is able to learn more effective common representations for different modalities in the shared subspace by fully excluding the interference within each modality. Extensive experiments conducted on cross-modal retrieval verify the advantages of our P3S method compared with 15 state-of-the-art methods on four widely used cross-modal datasets.

Cross-Modal Surface Material Retrieval Using Discriminant Adversarial Learning

The surface properties of an object play a vital role in the tasks of robotic manipulation or interaction with its surrounding environment. Tactile sensing can provide rich information about the surface properties of an object through physical contact. Hence, how to convey and interpret the tactile information to the user is a significant problem during the human–machine interaction. To this end, a visual–tactile cross-modal retrieval framework is proposed for perceptual estimation by associating tactile information to visual information of material surfaces. Namely, we can use tactile information of an unknown material surface to retrieve perceptually similar surfaces from an available surface visual sample set. For the proposed framework, we develop a discriminant adversarial learning method, which incorporates intramodal discriminant, cross-modal correlation, and intermodal consistency into a deep learning network for common feature representation learning. Experimental results on the publicly available data set show that the proposed framework and the method are effective.

Task-Driven Common Representation Learning via Bridge Neural Network

This paper introduces a novel deep learning based method, named bridge neural network (BNN) to dig the potential relationship between two given data sources task by task. The proposed approach employs two convolutional neural networks that project the two data sources into a feature space to learn the desired common representation required by the specific task. The training objective with artificial negative samples is introduced with the ability of mini-batch training and it’s asymptotically equivalent to maximizing the total correlation of the two data sources, which is verified by the theoretical analysis. The experiments on the tasks, including pair matching, canonical correlation analysis, transfer learning, and reconstruction demonstrate the state-of-the-art performance of BNN, which may provide new insights into the aspect of common representation learning.

Representation learning using step-based deep multi-modal autoencoders

Deep learning techniques have been successfully used in learning a common representation for multi-view data, wherein different modalities are projected onto a common subspace. In a broader perspective, the techniques used to investigate common representation learning falls under the categories of ‘canonical correlation-based’ approaches and ‘autoencoder-based’ approaches. In this paper, we investigate the performance of deep autoencoder-based methods on multi-view data. We propose a novel step-based correlation multi-modal deep convolution neural network (CorrMCNN) which reconstructs one view of the data given the other while increasing the interaction between the representations at each hidden layer or every intermediate step. The idea of step reconstruction reduces the constraint of reconstruction of original data, instead, the objective function is optimized for reconstruction of representative features. This helps the proposed model to generalize for representation and transfer learning tasks efficiently for high dimensional data. Finally, we evaluate the performance of the proposed model on three multi-view and cross-modal problems viz., audio articulation, cross-modal image retrieval and multilingual (cross-language) document classification. Through extensive experiments, we find that the proposed model performs much better than the current state-of-the-art deep learning techniques on all three multi-view and cross-modal tasks.

Deep Memory Network for Cross-Modal Retrieval

With the explosive growth of multimedia data on the Internet, cross-modal retrieval has attracted a great deal of attention in both computer vision and multimedia communities. However, this task is challenging due to the heterogeneity gap between different modalities. Current approaches typically involve a common representation learning process that maps data from different modalities into a common space by linear or nonlinear embedding. Yet, most of them only handle the dual-modal situation and generalize poorly to complex cases that involve multiple modalities. In addition, they often require expensive fine-grained alignment of training data among diverse modalities. In this paper, we address these with a novel cross-modal memory network (CMMN), in which memory contents across modalities are simultaneously learned from end to end without the need of exact alignment. We further account for the diversity across multiple modalities using the strategy of adversarial learning. Extensive experimental results on several large-scale datasets demonstrate that the proposed CMMN approach achieves state-of-the-art performance in the task of cross-modal retrieval.

CM-GANs

It is known that the inconsistent distributions and representations of different modalities, such as image and text, cause the heterogeneity gap, which makes it very challenging to correlate heterogeneous data and measure their similarities. Recently, generative adversarial networks (GANs) have been proposed and have shown their strong ability to model data distribution and learn discriminative representation. It has also been shown that adversarial learning can be fully exploited to learn discriminative common representations for bridging the heterogeneity gap. Inspired by this, we aim to effectively correlate large-scale heterogeneous data of different modalities with the power of GANs to model cross-modal joint distribution. In this article, we propose Cross-modal Generative Adversarial Networks (CM-GANs) with the following contributions. First, a cross-modal GAN architecture is proposed to model joint distribution over the data of different modalities. The inter-modality and intra-modality correlation can be explored simultaneously in generative and discriminative models. Both compete with each other to promote cross-modal correlation learning. Second, the cross-modal convolutional autoencoders with weight-sharing constraint are proposed to form the generative model. They not only exploit the cross-modal correlation for learning the common representations but also preserve reconstruction information for capturing the semantic consistency within each modality. Third, a cross-modal adversarial training mechanism is proposed, which uses two kinds of discriminative models to simultaneously conduct intra-modality and inter-modality discrimination. They can mutually boost to make the generated common representations more discriminative by the adversarial training process. In summary, our proposed CM-GAN approach can use GANs to perform cross-modal common representation learning by which the heterogeneous data can be effectively correlated. Extensive experiments are conducted to verify the performance of CM-GANs on cross-modal retrieval compared with 13 state-of-the-art methods on 4 cross-modal datasets.

Semi-supervised cross-modal common representation learning with vector-valued manifold regularization

While cross-media data, like text, image, audio, video and 3D model, has been the main form of big data, there is a current dearth of research on cross-media retrieval. In this paper, we focus on how to learn the common representation of heterogeneous data which is a key challenge for cross-media retrieval. Most existing approaches linearly project original low-level feature into a joint feature space for isomorphic data representation. However, linear projection cannot capture most complex cross-modal correlation with high nonlinearity. In this paper, we propose a novel feature learning algorithm, which is semi-supervised cross-modal vector-valued manifold regularization (SCVM), to explore common representation of heterogeneous data. SCVM jointly explores low-level feature correlation and semantic information in a unified framework. Based on manifold regularization, we learn cross-media features from vector-valued reproducing kernel Hilbert spaces (RKHS) by kernel transformation on both labeled and unlabeled samples; moreover, we impose smoothness constraints of possible solutions to improve retrieval accuracy. Comparing with the current state-of-the-art approaches on two public datasets, comprehensive experimental results show superior performance of our SCVM. The method is more robust and stable when extended from two media types to five media types, which is very attractive in practical application.