Multi-view Representation Learning Research Articles

Deep Multi-view Representation Learning for Video Anomaly Detection Using Spatiotemporal Autoencoders

Visual perception is a transformative technology that can recognize patterns from environments through visual inputs. Automatic surveillance of human activities has gained significant importance in both public and private spaces. It is often difficult to understand the complex dynamics of events in real-time scenarios due to camera movements, cluttered backgrounds, and occlusion. Existing anomaly detection systems are not efficient because of high intra-class variations and inter-class similarities existing among activities. Hence, there is a demand to explore different kinds of information extracted from surveillance videos to improve overall performance. This can be achieved by learning features from multiple forms (views) of the given raw input data. We propose two novel methods based on the multi-view representation learning framework. The first approach is a hybrid multi-view representation learning that combines deep features extracted from 3D spatiotemporal autoencoder (3D-STAE) and robust handcrafted features based on spatiotemporal autocorrelation of gradients. The second approach is a deep multi-view representation learning that combines deep features extracted from two-stream STAEs to detect anomalies. Results on three standard benchmark datasets, namely Avenue, Live Videos, and BEHAVE, show that the proposed multi-view representations modeled with one-class SVM perform significantly better than most of the recent state-of-the-art methods.

Unsupervised multi-view representation learning with proximity guided representation and generalized canonical correlation analysis

Multi-view data can collaborate with each other to provide more comprehensive information than single-view data. Although there exist a few unsupervised multi-view representation learning methods taking both the discrepancies and incorporating complementary information from different views into consideration, they always ignore the use of inner-view discriminant information. It remains challenging to learn a meaningful shared representation of multiple views. To overcome this difficulty, this paper proposes a novel unsupervised multi-view representation learning model, MRL. Unlike most state-of-art multi-view representation learning, which only can be used for clustering or classification task, our method explores the proximity guided representation from inner-view and complete the task of multi-label classification and clustering by the discrimination fusion representation simultaneously. MRL consists of three parts. The first part is a deep representation learning for each view and then aims to represent the latent specific discriminant characteristic of each view, the second part builds a proximity guided dynamic routing to preserve its inner features of direction,location and etc. At last, the third part, GCCA-based fusion, exploits the maximum correlations among multiple views based on Generalized Canonical Correlation Analysis (GCCA). To the best of our knowledge, the proposed MRL could be one of the first unsupervised multi-view representation learning models that work in proximity guided dynamic routing and GCCA modes. The proposed model MRL is tested on five multi-view datasets for two different tasks. In the task of multi-label classification, the results show that our model is superior to the state-of-the-art multi-view learning methods in precision, recall, F1 and accuracy. In clustering task, its performance is better than the latest related popular algorithms. And the performance varies w.r.t. the dimensionality of G is also made to explore the characteristics of MRL.

Few-shot Food Recognition via Multi-view Representation Learning

This article considers the problem of few-shot learning for food recognition. Automatic food recognition can support various applications, e.g., dietary assessment and food journaling. Most existing works focus on food recognition with large numbers of labelled samples, and fail to recognize food categories with few samples. To address this problem, we propose a Multi-View Few-Shot Learning (MVFSL) framework to explore additional ingredient information for few-shot food recognition. Besides category-oriented deep visual features, we introduce ingredient-supervised deep network to extract ingredient-oriented features. As general and intermediate attributes of food, ingredient-oriented features are informative and complementary to category-oriented features, and thus they play an important role in improving food recognition. Particularly in few-shot food recognition, ingredient information can bridge the gap between disjoint training categories and test categories. To take advantage of ingredient information, we fuse these two kinds of features by first combining their feature maps from their respective deep networks and then convolving combined feature maps. Such convolution is further incorporated into a multi-view relation network, which is capable of comparing pairwise images to enable fine-grained feature learning. MVFSL is trained in an end-to-end fashion for joint optimization on two types of feature learning subnetworks and relation subnetworks. Extensive experiments on different food datasets have consistently demonstrated the advantage of MVFSL in multi-view feature fusion. Furthermore, we extend another two types of networks, namely, Siamese Network and Matching Network, by introducing ingredient information for few-shot food recognition. Experimental results have also shown that introducing ingredient information into these two networks can improve the performance of few-shot food recognition.

Multi-View Representation Learning With Deep Gaussian Processes.

Multi-view representation learning is a promising and challenging research topic, which aims to integrate multiple data information from different views to improve the learning performance. The recent deep Gaussian processes (DGPs) have the advantages of good uncertainty estimates, powerful non-linear mapping ability and great generalization capability, which can be used as an excellent data representation learning method. However, DGPs only focus on single view data and are rarely applied to the multi-view scenario. In this paper, we propose a multi-view representation learning algorithm with deep Gaussian processes (named MvDGPs), which inherits the advantages of deep Gaussian processes and multi-view representation learning, and can learn more effective representation of multi-view data. The MvDGPs consist of two stages. The first stage is multi-view data representation learning, which is mainly used to learn more comprehensive representations of multi-view data. The second stage is classifier design, which aims to select an appropriate classifier to better employ the representations obtained in the first stage. In contrast with DGPs, MvDGPs support asymmetrical modeling depths for different views of data, resulting in better characterizations of the discrepancies among different views. Experimental results on real-world multi-view data sets verify the effectiveness of the proposed algorithm, which indicates that MvDGPs can integrate the complementary information in multiple views to discover a good representation of the data.

Hashtag our stories: Hashtag recommendation for micro-videos via harnessing multiple modalities

Due to the short attention span and instant gratification phenomenon, micro-videos are growing exponentially while gaining more and more concerns. Yet the sheer number of micro-videos leads to severe information overload issues, making it difficult for users to identify their desired micro-videos. The hashtag, mainly utilized in the domain of the microblog or the image, is the indicator or the core idea of the target content and can be applied to various information retrieval scenarios (e.g., search, browse, and categorization). So far, however, little attention has been paid to perform the hashtag recommendation for micro-videos via harnessing multiple modalities.In this article, we devise a neural network-based solution, LOGO (short for “muLti-mOdal-based hashtaG recOmmendation”), to recommend hashtags for micro-videos by utilizing multiple modalities. The proposed LOGO approach first represents each modality as the combination of sequential units in it, weighted by the attention mechanism. In this way, the sequential and attentive features are captured simultaneously. After that, the LOGO integrates the representations of all modalities via a multi-view representation learning framework, which projects the representations into a common space under the restriction of the modality similarity. Ultimately, the LOGO feed the projections of three modalities in the common space and the embeddings of hashtags into a customized neural collaborative filtering framework to perform the hashtag recommendation. Extensive experiments on the scope of both overall performance comparison and micro-level analyses have well-justified the effectiveness and rationality of our proposed approach.

Diagnosis of Coronavirus Disease 2019 (COVID-19) With Structured Latent Multi-View Representation Learning.

Recently, the outbreak of Coronavirus Disease 2019 (COVID-19) has spread rapidly across the world. Due to the large number of infected patients and heavy labor for doctors, computer-aided diagnosis with machine learning algorithm is urgently needed, and could largely reduce the efforts of clinicians and accelerate the diagnosis process. Chest computed tomography (CT) has been recognized as an informative tool for diagnosis of the disease. In this study, we propose to conduct the diagnosis of COVID-19 with a series of features extracted from CT images. To fully explore multiple features describing CT images from different views, a unified latent representation is learned which can completely encode information from different aspects of features and is endowed with promising class structure for separability. Specifically, the completeness is guaranteed with a group of backward neural networks (each for one type of features), while by using class labels the representation is enforced to be compact within COVID-19/community-acquired pneumonia (CAP) and also a large margin is guaranteed between different types of pneumonia. In this way, our model can well avoid overfitting compared to the case of directly projecting high-dimensional features into classes. Extensive experimental results show that the proposed method outperforms all comparison methods, and rather stable performances are observed when varying the number of training data.

Attentive multi-view reinforcement learning

The reinforcement learning process usually takes millions of steps from scratch, due to the limited observation experience. More precisely, the representation approximated by a single deep network is usually limited for reinforcement learning agents. In this paper, we propose a novel multi-view deep attention network (MvDAN), which introduces multi-view representation learning into the reinforcement learning framework for the first time. Based on the multi-view scheme of function approximation, the proposed model approximates multiple view-specific policy or value functions in parallel by estimating the middle-level representation and integrates these functions based on attention mechanisms to generate a comprehensive strategy. Furthermore, we develop the multi-view generalized policy improvement to jointly optimize all policies instead of a single one. Compared with the single-view function approximation scheme in reinforcement learning methods, experimental results on eight Atari benchmarks show that MvDAN outperforms the state-of-the-art methods and has faster convergence and training stability.

Robust Multi-View Representation Learning (Student Abstract)

Multi-view data has become ubiquitous, especially with multi-sensor systems like self-driving cars or medical patient-side monitors. We propose two methods to approach robust multi-view representation learning with the aim of leveraging local relationships between views.The first is an extension of Canonical Correlation Analysis (CCA) where we consider multiple one-vs-rest CCA problems, one for each view. We use a group-sparsity penalty to encourage finding local relationships. The second method is a straightforward extension of a multi-view AutoEncoder with view-level drop-out.We demonstrate the effectiveness of these methods in simple synthetic experiments. We also describe heuristics and extensions to improve and/or expand on these methods.

Multi-View Deep Attention Network for Reinforcement Learning (Student Abstract)

The representation approximated by a single deep network is usually limited for reinforcement learning agents. We propose a novel multi-view deep attention network (MvDAN), which introduces multi-view representation learning into the reinforcement learning task for the first time. The proposed model approximates a set of strategies from multiple representations and combines these strategies based on attention mechanisms to provide a comprehensive strategy for a single-agent. Experimental results on eight Atari video games show that the MvDAN has effective competitive performance than single-view reinforcement learning methods.

Semi-Supervised Multi-View Deep Discriminant Representation Learning.

Learning an expressive representation from multi-view data is a key step in various real-world applications. In this paper, we propose a semi-supervised multi-view deep discriminant representation learning (SMDDRL) approach. Unlike existing joint or alignment multi-view representation learning methods that cannot simultaneously utilize the consensus and complementary properties of multi-view data to learn inter-view shared and intra-view specific representations, SMDDRL comprehensively exploits the consensus and complementary properties as well as learns both shared and specific representations by employing the shared and specific representation learning network. Unlike existing shared and specific multi-view representation learning methods that ignore the redundancy problem in representation learning, SMDDRL incorporates the orthogonality and adversarial similarity constraints to reduce the redundancy of learned representations. Moreover, to exploit the information contained in unlabeled data, we design a semi-supervised learning framework by combining deep metric learning and density clustering. Experimental results on three typical multi-view learning tasks, i.e., webpage classification, image classification, and document classification demonstrate the effectiveness of the proposed approach.

Marine Vessel Re-Identification: A Large-Scale Dataset and Global-and-Local Fusion-Based Discriminative Feature Learning

A marine vessel re-identification system has to determine whether or not different images represent the same vessel. Accurate vessel re-identification improves onshore closed-circuit television monitoring in a vessel traffic services system as well as onboard surveillance of surrounding vessels. However, because ships are rigid bodies and the marine environment is harsh, the accurate re-identification of vessels at sea can be very difficult. We describe a marine vessel-re-identification framework, Global-and-Local Fusion-based Multi-view Feature Learning (GLF-MVFL), which is based on a combination of global and fine-grained local features. GLF-MVFL combines cross-entropy loss with our newly-developed orientation-guided quintuplet loss. We exploit intrinsic features of marine vessels to optimize multi-view representation learning for re-identification. GLF-MVFL uses ResNet-50 as the backbone network to extract features for simultaneous quintuple input. It detects and discriminates between features and estimates viewpoints to form a comprehensive re-identification framework. We created an annotated large-scale vessel retrieval dataset, VesselID-539, which contains images from viewpoints similar to those of an autonomous surface vessel, to use in evaluating the performance of the model. Extensive experiments and analysis of the results obtained from using VesselID-539 demonstrate that our approach significantly increases the accuracy of vessel re-identification and is more effective and robust for images from different viewpoints than other approaches.

SkeletonNet: A Hybrid Network With a Skeleton-Embedding Process for Multi-View Image Representation Learning

Multi-view representation learning plays a fundamental role in multimedia data analysis. Some specific inter-view alignment principles are adopted in conventional models, where there is an assumption that different views share a common latent subspace. However, when dealing views on diverse semantic levels, the view-specific characteristics are neglected, and the divergent inconsistency of similarity measurements hinders sufficient information sharing. This paper proposes a hybrid deep network by introducing tensor factorization into the multi-view deep auto-encoder. The network adopts skeleton-embedding process for unsupervised multi-view subspace learning. It takes full consideration of view-specific characteristics, and leverages the strength of both shallow and deep architectures for modeling low- and high-level views, respectively. We first formulate the high-level-view semantic distribution as the underlying skeleton structure of the learned subspace, and then infer the local tangent structures according to the affinity propagation of low-level-view geometric correlations. As a consequence, more discriminative subspace representation can be learned from global semantic pivots to local geometric details. Experimental comparisons on three benchmark image datasets show the promising performance and flexibility of our model.

Multi-view representation learning in multi-task scene

Over recent decades have witnessed considerable progress in whether multi-task learning or multi-view learning, but the situation that consider both learning scenes simultaneously has received not too much attention. How to utilize multiple views latent representation of each single task to improve each learning task performance is a challenge problem. Based on this, we proposed a novel semi-supervised algorithm, termed as Multi-Task Multi-View learning based on Common and Special Features (MTMVCSF). In general, multi-views are the different aspects of an object and every view includes the underlying common or special information of this object. As a consequence, we will mine multiple views jointly latent factor of each learning task which consists of each view special feature and the common feature of all views. By this way, the original multi-task multi-view data has degenerated into multi-task data, and exploring the correlations among multiple tasks enables to make an improvement on the performance of learning algorithm. Another obvious advantage of this approach is that we get latent representation of the set of unlabeled instances by the constraint of regression task with labeled instances. The performance of classification and semi-supervised clustering task in these latent representations perform obviously better than it in raw data. Furthermore, an anti-noise multi-task multi-view algorithm called AN-MTMVCSF is proposed, which has a strong adaptability to noise labels. The effectiveness of these algorithms is proved by a series of well-designed experiments on both real world and synthetic data.

Multi-View Saliency Guided Deep Neural Network for 3-D Object Retrieval and Classification

In this paper, we propose the multi-view saliency guided deep neural network (MVSG-DNN) for 3D object retrieval and classification. This method mainly consists of three key modules. First, the module of model projection rendering is employed to capture the multiple views of one 3D object. Second, the module of visual context learning applies the basic Convolutional Neural Networks for visual feature extraction of individual views and then employs the saliency LSTM to adaptively select the representative views based on multi-view context. Finally, with these information, the module of multi-view representation learning can generate the compile 3D object descriptors with the designed classification LSTM for 3D object retrieval and classification. The proposed MVSG-DNN has two main contributions: 1) It can jointly realize the selection of representative views and the similarity measure by fully exploiting multi-view context; 2) It can discover the discriminative structure of multi-view sequence without constraints of specific camera settings. Consequently, it can support flexible 3D object retrieval and classification for real applications by avoiding the required camera settings. Extensive comparison experiments on ModelNet10, ModelNet40, and ShapeNetCore55 demonstrate the superiority of MVSG-DNN against the state-of-art methods.

Multi-view representation learning for segmentation of abnormal tissues in medical images applied to multiple sclerosis lesion delineation

Automated segmentation of abnormal tissues in medical images assists both physicians and medical researchers in the process of diseases diagnostic and research activities respectively. Intelligent techniques of automated segmentation are gaining more popularity in contrast to non-intelligent ones. In these techniques, quality representation of pixel/voxels by considering multiple natural and artificial views which exist in medical images increases segmentation accuracy. The proposed method for segmentation of abnormal tissues in medical images is based on multi-view representation with six phases of pre-processing, view generation, representation generation, classification, post-processing, and evaluation. In the representation phase, raw data of medical images are represented based on the modes of variation or clusters exist in the original multi-view feature space. Quantitative results of the experiment demonstrate representations generated via the proposed method are effective especially when the Random Forest classifier is employed. DSC of 0.72 for a subject shows that the results are promising. This study shows cluster based representation of raw pixel/voxels of multiple views are effective in supervised segmentation of abnormal tissues.

Letting Logos Speak: Leveraging Multiview Representation Learning for Data-Driven Logo Design

Logos serve a fundamental role as the visual figureheads of brands. Yet, due to the difficulty of using unstructured image data, prior research on logo design has largely been limited to non-quantitative studies. In this work, we explore the interplay between logo design and brand identity creation from a data-driven perspective. We develop both a novel logo feature extraction algorithm that uses modern image processing tools to decompose pixel-level image data into meaningful features, and a multiview representation learning framework that links these visual features to textual descriptions, consumer ratings of brand personality, and other high-level tags describing firms. We apply this framework to a unique dataset of brands, to understand which brands use which logo features, and how consumers evaluate these brands’ personalities. Moreover, we show that manipulating the model’s learned representations through what we term “brand arithmetic” yields new brand identities, and can help with ideation. Finally, through an application to fast food branding, we show how our model can be used as a decision support tool for suggesting typical logo features for a brand, and for predicting consumers’ reactions to new brands or rebranding efforts.

Canonical Correlation Analysis With L2,1-Norm for Multiview Data Representation.

For many machine learning algorithms, their success heavily depends on data representation. In this paper, we present an l2,1-norm constrained canonical correlation analysis (CCA) model, that is, L2,1-CCA, toward discovering compact and discriminative representation for the data associated with multiple views. To well exploit the complementary and coherent information across multiple views, the l2,1-norm is employed to constrain the canonical loadings and measure the canonical correlation loss term simultaneously. It enables, on the one hand, the canonical loadings to be with the capacity of variable selection for facilitating the interpretability of the learned canonical variables, and on the other hand, the learned canonical common representation keeps highly consistent with the most canonical variables from each view of the data. Meanwhile, the proposed L2,1-CCA can also be provided with the desired insensitivity to noise (outliers) to some degree. To solve the optimization problem, we develop an efficient alternating optimization algorithm and give its convergence analysis both theoretically and experimentally. Considerable experiment results on several real-world datasets have demonstrated that L2,1-CCA can achieve competitive or better performance in comparison with some representative approaches for multiview representation learning.

Multi-View Network Representation Learning Algorithm Research

Network representation learning is a key research field in network data mining. In this paper, we propose a novel multi-view network representation algorithm (MVNR), which embeds multi-scale relations of network vertices into the low dimensional representation space. In contrast to existing approaches, MVNR explicitly encodes higher order information using k-step networks. In addition, we introduce the matrix forest index as a kind of network feature, which can be applied to balance the representation weights of different network views. We also research the relevance amongst MVNR and several excellent research achievements, including DeepWalk, node2vec and GraRep and so forth. We conduct our experiment on several real-world citation datasets and demonstrate that MVNR outperforms some new approaches using neural matrix factorization. Specifically, we demonstrate the efficiency of MVNR on network classification, visualization and link prediction tasks.

Adversarial correlated autoencoder for unsupervised multi-view representation learning

To eliminate the view discrepancy of multi-view data due to different distributions, the key is to learn the common representation for multi-view data in many practical applications. To achieve the end, we propose a novel unsupervised multi-view representation learning method (called Adversarial Correlated AutoEncoder, AdvCAE). In brief, AdvCAE utilizes a deep structure to achieve nonlinear representation and adversarial learning scheme for distribution matching. To be specific, AdvCAE performs like an adversarial autoencoder (AAE) which could conduct variational inference by matching the aggregated posteriors of the latent variable with a specific prior distribution. Benefiting from our model, the representations of different views could follow the same distribution, thus learning the common representation for different views. To the best of our knowledge, AdvCAE could be one of the first unsupervised multi-view representation learning approaches that work in the manner of adversarial learning. To verify the effectiveness of the proposed method, we conduct experiments on five public real-world datasets w.r.t. the applications of cross-view classification and cross-view retrieval tasks. The experimental results show that our method remarkably outperforms than 15 state-of-the-art methods.

A Survey of Multi-View Representation Learning

Recently, multi-view representation learning has become a rapidly growing direction in machine learning and data mining areas. This paper introduces two categories for multi-view representation learning: multi-view representation alignment and multi-view representation fusion. Consequently, we first review the representative methods and theories of multi-view representation learning based on the perspective of alignment, such as correlation-based alignment. Representative examples are canonical correlation analysis (CCA) and its several extensions. Then from the perspective of representation fusion we investigate the advancement of multi-view representation learning that ranges from generative methods including multi-modal topic learning, multi-view sparse coding, and multi-view latent space Markov networks, to neural network-based methods including multi-modal autoencoders, multi-view convolutional neural networks, and multi-modal recurrent neural networks. Further, we also investigate several important applications of multi-view representation learning. Overall, this survey aims to provide an insightful overview of theoretical foundation and state-of-the-art developments in the field of multi-view representation learning and to help researchers find the most appropriate tools for particular applications.