Search

ComptoxAI is a new data infrastructure for computational and artificial intelligence research in predictive toxicology. Here, we describe and showcase ComptoxAI's graph-structured knowledge base in the context of three real-world use-cases, demonstrating that it can rapidly answer complex questions about toxicology that are infeasible using previous technologies and data resources. These use-cases each demonstrate a tool for information retrieval from the knowledge base being used to solve a specific task: The "shortest path" module is used to identify mechanistic links between perfluorooctanoic acid (PFOA) exposure and nonalcoholic fatty liver disease; the "expand network" module identifies communities that are linked to dioxin toxicity; and the quantitative structure-activity relationship (QSAR) dataset generator predicts pregnane X receptor agonism in a set of 4,021 pesticide ingredients. The contents of ComptoxAI's source data are rigorously aggregated from a diverse array of public third-party databases, and ComptoxAI is designed as a free, public, and open-source toolkit to enable diverse classes of users including biomedical researchers, public health and regulatory officials, and the general public to predict toxicology of unknowns and modes of action.

Analyzing short texts infers discriminative and coherent latent topics that is a critical and fundamental task since many real-world applications require semantic understanding of short texts. Traditional long text topic modeling algorithms (e.g., PLSA and LDA) based on word co-occurrences cannot solve this problem very well since only very limited word co-occurrence information is available in short texts. Therefore, short text topic modeling has already attracted much attention from the machine learning research community in recent years, which aims at overcoming the problem of sparseness in short texts. In this survey, we conduct a comprehensive review of various short text topic modeling techniques proposed in the literature. We present three categories of methods based on Dirichlet multinomial mixture, global word co-occurrences, and self-aggregation, with example of representative approaches in each category and analysis of their performance on various tasks. We develop the first comprehensive open-source library, called STTM, for use in Java that integrates all surveyed algorithms within a unified interface, benchmark datasets, to facilitate the expansion of new methods in this research field. Finally, we evaluate these state-of-the-art methods on many real-world datasets and compare their performance against one another and versus long text topic modeling algorithm., arXiv admin note: text overlap with arXiv:1808.02215 by other authors

Cross-view hashing has shown great potential for large-scale retrieval due to its superiority in terms of computation and storage. In real-world applications, data emerges in a streaming manner, e.g., new images and tags are uploaded to social media by users every day. Existing cross-view hashing methods have to retrain model on new multi-view data, which is time-consuming and not applicable in the real-world applications. To fill this gap, this paper proposes a new online cross-view hashing method, dubbed online unsupervised cross-view discrete hashing (OUCDH) that considers similarity preservation and quantization loss. OUCDH generates hash code as latent embedding shared by multiple views via matrix factorization. OUCDH can well preserve similarity among newly arriving data and old data with the help of anchor graph. An iterative efficient algorithm is developed for online optimization. OUCDH further updates hash code of old data to match that of newly arriving data in each iteration. Extensive experiments on three benchmarks demonstrate that the proposed OUCDH yields superior performance than existing state-of-the-art online cross-view hashing methods.

How to represent and classify a testing sample for the representation-based classification (RBC) plays an important role in the filed of pattern recognition. As a typical kind of the representation-based classification with promising performance, collaborative representation-based classification (CRC) adopts all the training samples to collaboratively represent and then classify each testing sample with the reconstructive residuals among all the classes. However, most of the CRC methods fail to make full use of the localities and discrimination information of data in collaborative representation. To address this issue to further improve the classification performance, we design a novel supervised CRC method entitled locality-constrained weighted collaborative-competitive representation-based classification (LWCCRC). In the proposed method, the localities of data are taken into account by using the positive and negative nearest samples of each testing sample with their corresponding weighted constraints. Such devised locality-constrained weighted term can model the similarity and natural discrimination information contained in the neighborhood region for each testing sample to obtain the favorable representation. Moreover, a competitive constraint is introduced to enhance pattern discrimination among the categorical collaborative representations. To explore the effectiveness of our proposed LWCCRC, the extensive experiments are carried out on three different types of data sets. The experimental results demonstrate that the proposed LWCCRC significantly outperforms the recent state-of-the-art CRC methods.

Face super-resolution (FSR) is an effective way to deal with low-resolution (LR) face images, which can infer the latent high-resolution (HR) face images from the LR inputs. In contrast with traditional FSR methods such as interpolation, learning-based methods generate more realistic HR images of LR faces by exploiting the relationship between HR and LR images . In this paper, we propose a novel FSR learning approach based on orthonormalized partial least squares referred to as OPLS-SR. It first learns a latent coherent feature space of low-dimensional HR and LR face embeddings via a recursive optimization, and then super-resolves the LR face images through global face reconstruction and facial detail compensation. Experimental results on the CAS-PEAL-R1 and FERET face databases have demonstrated the effectiveness of the proposed OPLS-SR method in terms of quantitative and qualitative evaluations.

Simultaneous localization and object tracking (SLOT) is essentially important for autonomous systems. Tightly-coupled and loosely-coupled methods are two commonly used back-end frameworks for the state-of-the-art solutions of SLOT problem. However, some inherent limitations exist in these two frameworks. In particular, the tightly-coupled method is usually disturbed by the poor observations of some dynamic objects, and the performance of a loosely-coupled one completely depends on that of classical static simultaneous localization and mapping (SLAM) process. Motivated by these observations, we propose a novel switching-coupled back-end solution and theoretically derive its concrete form using probability representation. Based on the switching strategy and the proposed objects classification criteria where the object uncertainty, observation quality and prior information are jointly considered, the dynamic objects’ states are flexibly coupled with camera's state and static landmarks’ states. For implementation, the measurement constraints of “good” dynamic objects and static landmarks are simultaneously leveraged to perform SLAM and good object tracking (SLAMGOT) process, and those of “bad” objects are used for bad object tracking (BOT) process based on the obtained camera state. Extensive evaluations on synthetic scenes, KITTI datasets and real-world experiments demonstrate the performance of the proposed method.

Depending on the scanning mode, existing short text stream clustering methods can be divided into the following two kinds of methods: one-pass-based and batch-based. The one-pass-based method handles each text only one time, but cannot deal with the sparseness problem very well. The batch-based method obtains better results by allowing multiple iterations of each batch, but the efficiency is relatively low. To overcome these problems, this paper presents Lifelong learning Augmented Short Text stream clustering method (LAST), which incorporates the episodic memory module and sparse experience replay module of lifelong learning into the clustering process. Specifically, LAST processes each text one time, but at a certain interval it randomly samples some previously seen texts of the episodic memory to update cluster features by performing sparse experience replay. Empirical studies on two public datasets demonstrate that the performance of the LAST-based method is on a par with the batch-based method, and runs close to the speed of the one-pass-based method.

Multi-label classification aims to deal with the problem that an object may be associated with one or more labels, which is a more difficult task due to the complex nature of multi-label data. The crucial problem of multi-label classification is the more robust and higher-level feature representation learning, which can reduce non-helpful feature attributes from the input space prior to training. In recent years, deep learning methods based on autoencoders have achieved excellent performance in multi-label classification for the advantages of powerful representations learning ability and fast convergence speed. However, most existing autoencoder-based methods only rely on the single autoencoder model, which pose challenges for multi-label feature representations learning and fail to measure similarities between data spaces. To address this problem, in this paper, we propose a novel representation learning method with dual autoencoder for multi-label classification. Compared to the existing autoencoder-based methods, our proposed method can capture different characteristics and more abstract features from data by the serially connection of two different types of autoencoders. More specifically, firstly, the algorithm of Reconstruction Independent Component Analysis (RICA) in sparse autoencoder is trained on patches on all training and test dataset for robust global feature representations learning. Secondly, with the output of RICA, stacked autoencoder with manifold regularization (SAMR) is introduced to ameliorate the quality of multi-label features learning. Comprehensive experiments on several real-world data sets demonstrate the effectiveness of our proposed approach compared with several competing state-of-the-art methods.

Lexical simplification has attracted much attention in many languages, which is the process of replacing complex words in a given sentence with simpler alternatives of equivalent meaning. Although the richness of vocabulary in Chinese makes the text very difficult to read for children and non-native speakers, there is no research work for the Chinese lexical simplification (CLS) task. To circumvent difficulties in acquiring annotations, we manually create the first benchmark dataset for CLS, which can be used for evaluating the lexical simplification systems automatically. To acquire a more thorough comparison, we present five different types of methods as baselines to generate substitute candidates for the complex word that includes synonym-based approach, word embedding-based approach, BERT-based approach, sememe-based approach, and a hybrid approach. Finally, we design the experimental evaluation of these baselines and discuss their advantages and disadvantages. To our best knowledge, this is the first study for CLS task.

Deep hashing has greatly improved retrieval performance with the powerful learning capability of deep neural network. However, deep unsupervised hashing can hardly achieve impressive performance due to the lack of the semantic supervision. This letter proposes Unsupervised Discriminative Deep Hashing (UD $^2$ H) to fulfill this gap. UD $^2$ H is formulated to jointly perform hash code learning and clustering, and trained in an asymmetric manner to improve the efficiency. The cluster labels supervise the training of deep model to enable hash code discriminative. Based on the outputs of the deep model, UD $^2$ H adaptively constructs a similarity graph that considers the local and global structures. Experiments on three benchmark datasets show that the proposed UD $^2$ H outperforms the state-of-the-art unsupervised deep hashing methods.

Lexical simplification (LS) aims at replacing complex words with simpler alternatives. LS commonly consists of three main steps: complex word identification, substitute generation, and substitute ranking. Existing LS methods focus on the contextual information of the complex word in the last step (substitute ranking). However, they miss out the following two facts: (1) The word complexity of a polysemous word is very closely related to its context; (2) The step of substitute generation regardless of the context will inevitably produce a large number of spurious candidates. Therefore, we propose a novel LS system LSBert based on pretrained language model BERT to address the aforementioned issues, which is capable of making use of the wider context when both identifying the words in need of simplification and generating substitute candidates for the complex words.Specifically, LSBert consists of a network for complex word identification by fine-tuning BERT and a network for substitute generation based on BERT. Experimental results show that LSBert performs well in both complex word identification and substitute generation, achieving state-of-the-art results in three benchmarks. To facilitate reproducibility, the code of the LSBert system is available at https://github.com/qiang2100/BERT-LS .

A multi-sensor fusion approach for simultaneous localisation and mapping (SLAM) based on a bio-inspired polarised skylight sensor is presented in this study. The innovation of the proposed approach is that a newly designed bio-inspired polarised skylight sensor, which is inspired by the navigation principle of desert ant, is introduced to improve the accuracy of SLAM. The measurement equations based on a polarised skylight sensor and a lidar are derived to obtain the orientation and position of the mobile robot and landmarks. Three kinds of non-linear filters, extended Kalman filter (EKF), unscented Kalman filter, and particle filter, are adopted and compared to fuse the polarised skylight sensor, lidar, and odometry to estimate the position, orientation, and map in the experiments. Simulation tests and experiments are conducted to validate the effectiveness of the proposed method. The simulations show that the EKF-SLAM with the polarised skylight sensor reduces the error of localisation about 30% and the error of mapping about 25%. Experiments indicate that the proposed EKF-SLAM approach can reduce the error of position and the heading angle, which verifies the proposed EKF-SLAM method, can be used for the outdoor with the low-cost multi-sensor.

Vehicle re-identification (re-ID) plays an important role in the automatic analysis of the increasing urban surveillance videos and has become a hot topic in recent years. Vehicle re-ID aims at identifying vehicles across different cameras. However, it suffers from the difficulties caused by various viewpoint of vehicles, diversified illuminations, and complicated environments. In this paper, a two-stage vehicle re-ID framework is proposed to address these challenges, which contains a feature extraction module for achieving discriminative features and a spatial-temporal re-ranking module to improve the accuracy of vehicle re-ID task. Firstly, a multi-task deep network that integrates identity predicting network, attribute recognition network and verification network is adopted to learn discriminate features. Secondly, a spatio-temporal model is built to re-rank the appearance information measurement results, which utilizes the spatio-temporal relationship to increase constraints of the images. Moreover, to facilitate progressive vehicle re-ID research, experiments are conducted on both the VeRi-776 dataset and VehicleID dataset. Both the proposed multi-task feature extraction module and spatio-temporal model achieve considerable improvements.

Graph embedding has attracted much more research interests in dimensionality reduction. In this study, based on collaborative representation and graph embedding, the authors propose a new linear dimensionality reduction method called collaborative representation-based locality preserving projection (CRLPP). In the CRLPP, they assume that the similar samples should have similar reconstructions by collaborative representation and the similar reconstructions should also have the similar low-dimensional representations in the projected subspace. CRLPP first reconstructs each training sample using the collaborative representation of the other remaining training samples, and then designs the graph construction of all training samples, finally establishes the objective function of graph embedding using the collaborative reconstructions and the constructed graph. The proposed CRLPP can well preserve the intrinsic geometrical and discriminant structures of high-dimensional data in low-dimensional subspace. The effectiveness of the proposed is verified on several image datasets. The experimental results show that the proposed method outperforms the state-of-art dimensionality reduction.

Last decades have witnessed the significance development of research in short text classification. However, most existing methods only focus on the text which contained dozens of words like Twitter or MicroBlog, but not take the short text with few words like news headline or invoice name into consideration. Meanwhile, contemporary short text classification methods either to expand feature of short text with external corpus or to learn the feature representation from all the texts, which have not take the difference between words of short text into full consideration. Notably, the classification of short text with few words are usually determined by a few specific key words contrary to documents classification or traditional short text classification. To address these problems, this paper propose a hybrid classification method of Attention mechanism and Feature selection via Character embedding in Chinese short text with few words, called AFC. More specifically, firstly, the character embedding is computed to represent Chinese short texts with few words, which takes full advantage of short text information without external corpus. Secondly, attention-based LSTM is introduced in our method to project the data into feature representation space with weighting, which make the keywords in classification have more subtle value. Furthermore, the semantic similarity between content and class label information is calculated for feature selection, which reduces the possible negative influence of some redundant information on classification. Experiments on real-world datasets demonstrate the effectiveness of our method compared to other competing methods.

In multi-view learning, there are a large number of small sample size problems in which the number of training samples is less than the dimension of feature vectors. For such problems, generalized multi-view linear discriminant analysis (GMLDA) usually fails to achieve good learning performance for many classification tasks. In this paper, we propose a new multi-view feature extraction method via fractional spectral modeling, namely, fractional generalized multi-view discriminant analysis (FGMDA), which is able to subsume GMLDA as a special case. In addition, an uncorrelated FGMDA (UFGMDA) is further proposed, where the extracted features are mutually uncorrelated in each view and thus have minimum redundancy. Extensive experimental results on visual recognition demonstrate the superiority of our proposed methods over the state-of-the-art methods in terms of classification accuracy.

The probabilistic collaborative representation-based classification (PCRC), as a novel extension of collaborative representation-based classification (CRC), is a promising method in pattern recognition. In this article, we adopt the coarse to fine representation to propose two-phase probabilistic collaborative representation based-classification (TPCRC) to enhance the power of pattern discrimination in PCRC. In TPCRC, the first phase is to utilize probabilistic collaborative representation to coarsely choose the nearest representative samples, and the second phase is to use the chosen nearest samples to finely represent and classify each testing sample. In order to employ the locality of data to further improve classification performance of PCRC, we also propose two-phase weighted probabilistic collaborative representation based-classification (TWPCRC). In the fine representation of TWPCRC, the probabilistic collaborative coefficients are weighted by the local distance similarities between each testing sample and all the training samples. The proposed methods are verified by the comparative experiments on three public image data sets. Experimental results show that the proposed methods outperform the state-of-the-art collaborative representation-based classification methods.

Quantitative Structure-Activity Relationship (QSAR) modeling is the most common computational technique for predicting chemical toxicity, but a lack of methodological innovations in QSAR have led to underwhelming performance. We show that contemporary QSAR modeling for predictive toxicology can be substantially improved by incorporating semantic graph data aggregated from open-access public databases, and analyzing those data in the context of graph neural networks (GNNs). Furthermore, we introspect the GNNs to demonstrate how they can lead to more interpretable applications of QSAR, and use ablation analysis to explore the contribution of different data elements to the final models’ performance.

Self-supervised learning (SSL) has been proved very effective in learning representations from unlabeled data in language and vision domains. Yet, very few instrumental self-supervised approaches exist for 3D skeleton action understanding, and directly applying the existing SSL methods from other domains for skeleton action learning may suffer from misalignment of representations and some limitations. In this paper, we consider that a good representation learning encoder can distinguish the underlying features of different actions, which can make the similar motions closer while pushing the dissimilar motions away. There exists, however, some uncertainties in the skeleton actions due to the inherent ambiguity of 3D skeleton pose in different viewpoints or the sampling algorithm in contrastive learning, thus, it is ill-posed to differentiate the action features in the deterministic embedding space. To address these issues, we rethink the distance between action features and propose to model each action representation into the probabilistic embedding space to alleviate the uncertainties upon encountering the ambiguous 3D skeleton inputs. To validate the effectiveness of the proposed method, extensive experiments are conducted on Kinetics, NTU60, NTU120, and PKUMMD datasets with several alternative network architectures. Experimental evaluations demonstrate the superiority of our approach and through which, we can gain significant performance improvement without using extra labeled data.

In recent years, the demand for Wi-Fi has grown exponentially, which has led to the rapid development of indoor positioning services based on Wi-Fi fingerprints. Due to the Received Signal Strength Indicator (RSSI) variation over time, the device heterogeneity, and dynamic changes in the environment, the accuracy and robustness of traditional Wi-Fi fingerprint-based methods usually degrade. To alleviate these issues, we propose an approach based on the denoising supervised autoencoder for regression tasks that we named Wi-Fi DSAR. The idea of Wi-Fi DSAR design is to (a) has strong resistance to the noise in the received Wi-Fi signal, and (b) prevent overfitting the fingerprint database to achieve robustness. We do a performance study by comparing other fingerprint-based works with our Wi-Fi DSAR. All works are evaluated in three open datasets, i.e., DSI, IPIN2016, and IPIN2020, with different area sizes and access point densities. Experimental results demonstrate that compared with existing fingerprint-based methods, our Wi-Fi DSAR is able to reduce the average positioning error by 20% to 50%. This is because our Wi-Fi DSAR has both noise immunity and generalization performance.

Advanced Driver Assistance System (ADAS) and Artificial Intelligent (AI) are the important issue in recent years, autonomous car plays an important role in whole ADAS. To detect the environment surround the car, the sensor might be sensitive and immediate. LiDAR (Light Detection and Ranging) uses Laser to get the reflectivity from the surrounding objects. For clustering the objects with point cloud, the density of the point cloud still sparse, making the cluster result completely, we implement a system combines LiDAR and multi-view image, the depth image is generated by multi-view can help us to cluster the object in point cloud clearly. In addition, we use CORDIC (Coordinate Rotation Digital Computer) EEAS (Extended Elementary Angle Set) architecture to decode the package data that collected from Velodyne HDL-64E. By using the flow of digital chip design, we reduce the power consumption and accelerate the speed. The proposed system achieves 91.09% accuracy and the processing time is 0.757 second.

Weakly supervised object localization (WSOL) aims to localize objects with only image-level labels. Previous methods often try to utilize feature maps and classification weights to localize objects using image level annotations indirectly. In this paper, we demonstrate that weakly supervised object localization should be divided into two parts: class-agnostic object localization and object classification. For class-agnostic object localization, we should use class-agnostic methods to generate noisy pseudo annotations and then perform bounding box regression on them without class labels. We propose the pseudo supervised object localization (PSOL) method as a new way to solve WSOL. Our PSOL models have good transferability across different datasets without fine-tuning. With generated pseudo bounding boxes, we achieve 58.00% localization accuracy on ImageNet and 74.97% localization accuracy on CUB-200, which have a large edge over previous models., Accepted by CVPR 2020; Corrected some typo in the paper; The code repository is https://github.com/tzzcl/PSOL

Face hallucination (FH) is a powerful technique to reconstruct high-resolution (HR) faces from low-resolution (LR) faces. Most of conventional FH techniques ignore the influence of small training data, which may lead to the bias of variance and covariance. In this paper, we propose a novel FH method via fractional orthogonal latent consistent features that we call fractional orthogonal partial least squares based FH (FOPLSFH). In the proposed FOPLS-FH, intra- and cross-resolution covariance matrices are re-estimated through fractional-order eigenvalues and singular values modeling. Experimental results on real-world face datasets demonstrate the effectiveness of the proposed FOPLS-FH method.

Canonical correlation analysis (CCA) is a popular and powerful technique for two-view dimension reduction and feature extraction. But, CCA is not able to directly handle more than two view data and has a rigorous assumption that all the samples from two different views are paired. However, practical multiple view data are often semi-paired. To address this problem, we in this paper propose a novel semi-paired multiview dimension reduction approach, which takes cross-view neighborhood relationship among semi-paired data and within-view global structure information into consideration. The proposed approach can not only deal with multiview (more than two) data, but also take sufficient advantage of unpaired multiview data and then mitigate overfitting effectively caused by the limited paired data. Experimental results on two benchmark data sets demonstrate the effectiveness of our proposed method.

Hashing techniques have recently gained increasing research interest in multimedia studies. Most existing hashing methods only employ single features for hash code learning. Multiview data with each view corresponding to a type of feature generally provides more comprehensive information. How to efficiently integrate multiple views for learning compact hash codes still remains challenging. In this article, we propose a novel unsupervised hashing method, dubbed multiview discrete hashing (MvDH), by effectively exploring multiview data. Specifically, MvDH performs matrix factorization to generate the hash codes as the latent representations shared by multiple views, during which spectral clustering is performed simultaneously. The joint learning of hash codes and cluster labels enables that MvDH can generate more discriminative hash codes, which are optimal for classification. An efficient alternating algorithm is developed to solve the proposed optimization problem with guaranteed convergence and low computational complexity. The binary codes are optimized via the discrete cyclic coordinate descent (DCC) method to reduce the quantization errors. Extensive experimental results on three large-scale benchmark datasets demonstrate the superiority of the proposed method over several state-of-the-art methods in terms of both accuracy and scalability.

Recently many topic models such as Latent Dirichlet Allocation (LDA) and Non-negative Matrix Factorization (NMF) have made important progress towards generating high-level knowledge from a large corpus. However, these algorithms based on random initialization generate different results on the same corpus using the same parameters, denoted as instability problem. For solving this problem, ensembles of NMF are known to be much more stable and accurate than individual NMFs. However, training multiple NMFs for ensembling is computationally expensive. In this paper, we propose a novel scheme to obtain the seemingly contradictory goal of ensembling multiple NMFs without any additional training cost. We train a single NMF algorithm with the cyclical learning rate schedule, which can converge to several local minima along its optimization path. We save the results to the ensemble when the model converges, and then restart the optimization with a large learning rate that can help escape the current local minimum. Based on experiments performed on text corpora using a number of measures to assess, our method can reduce instability at no additional training cost, while simultaneously yields more accurate topic models than traditional single methods and ensemble methods.

Multi-view data with each view corresponding to a type of feature generally provides more comprehensive information. Learning from multi-view data is a challenging research topic in pattern recognition. For recognition task, most multi-view learning methods separately learn multi-view dimensionality reduction (MvDR) and classification models. Thus, the connection between the two models has not been well studied. In this paper, we propose a novel multi-view dimensionality reduction and recognition framework, which can establish the connection between MvDR and classification. Specifically, a multi-view dimensionality reduction method, termed as sparse representation regularized multiset canonical correlation analysis (SR2MCC) is first proposed. SR2MCC considers both correlation and sparse discrimination among multiple views. In accord with SR2MCC, a classifier, called multi-view sparse representation based classifier (MvSRC) is further developed. MvSRC performs classification by comparing the reconstruction residuals of different classes among all views. An efficient iterative algorithm is proposed to solve the proposed model. Extensive experiments on the AR, CMU PIE, FERET, and FRGC datasets demonstrate that the proposed framework can achieve superior recognition performance than several state-of-the-art methods.

Sparse representation models have been shown promising results for image denoising. However, conventional sparse representation-based models cannot obtain satisfactory estimations for sparse coefficients and the dictionary. To address this weakness, in this paper, we propose a novel fractional-order sparse representation U+0028 FSR U+0029 model. Specifically, we cluster the image patches into K groups, and calculate the singular values for each clean U+002F noisy patch pair in the wavelet domain. Then the uniform fractional-order parameters are learned for each cluster. Then a novel fractional-order sample space is constructed using adaptive fractional-order parameters in the wavelet domain to obtain more accurate sparse coefficients and dictionary for image denoising. Extensive experimental results show that the proposed model outperforms state-of-the-art sparse representation-based models and the block-matching and 3D filtering algorithm in terms of denoising performance and the computational efficiency.

In this paper, a novel LDA-based dimensionality reduction method called fractional-order embedding direct LDA (FEDLDA) is proposed. More specifically, we redefine the fractional-order between-class and within-class scatter matrices which can significantly reduce the deviation of sample covariance matrices caused by the noise disturbance and limited number of training samples; then the novel feature extraction criterion based on the direct LDA (DLDA) and the idea of fractional-order embedding is applied. Experiments on AT&T, Yale and AR face image databases are performed to test and evaluate the effectiveness of the proposed algorithms. Extensive experimental results show that FEDLDA outperforms DLDA and other closely related methods in terms of classification accuracy and efficiency.

As the representative one of representation-based classification (RBC) methods, collaborative RBC (CRC) has drawn much attention in pattern recognition and machine learning recently. Moreover, the collaborative representation-based face recognition has been extensively studied because of the effective classification performance of CRC. CRC collaboratively represents each query sample as the linear combination of all the training samples and then classifies the query sample according to the categorical representation-based distances. However, most variants of CRC cannot fully consider the locality and discrimination of data and cannot well handle the noise data, which has negative effect on real-world classification problems, such as face recognition. In this paper, a new discriminative collaborative neighbor representation (DCNR) method for face recognition is proposed by integrating class discrimination and data locality. In the proposed method, the locality of data constrains collaborative representation of each query sample to find representative nearest samples of the query sample. Moreover, the class discrimination regularization is taken into account by employing the representation of each class for each query sample. Due to the existing noises, such as corruptions and occlusions in face recognition, we further propose robust DCNR (R-DCNR) for robust classification by using the ℓ1-norm representation fidelity. Extensive experiments on face databases demonstrate that the proposed methods achieve competitive classification performance, compared to the state-of-the-art representation-based classification methods.

In the past decades, recommendation systems have provided lots of valuable personalized suggestions for the users to address the problem of information over-loaded. Collaborative Filtering (CF) is one of the most commonly applied and successful recommendation approaches, which refers to using the preferences of groups with similar interests to recommend information to other users. Recently, in addition to the traditional matrix factorization techniques, deep learning methods have been proposed to learn more abstract and higher-level representations for recommendation. However, most previous deep recommendation methods learn the higher-level feature representations of users and items through an identical model structure, which ignores the different characteristics of the user-based and item-based data. In addition, the rating matrix is usually sparse which may result in a significant degradation of recommendation performance. To address these problems, we propose a representation learning method with Collaborative Autoencoder for Personalized Recommendation (CAPR for short). In this method, user-based and item-based feature representations are learned by two different autoencoders for capturing different features of the data. Meanwhile, items’ attributions are combined into the feature representations with semi-autoencoder for alleviating the sparsity problem. Extensive experimental results confirm the effectiveness of our proposed method compared to other state-of-the-art matrix factorization methods and deep recommendation methods.

Due to the significant reduction in computational cost and storage, hashing techniques have gained increasing interests in facilitating large-scale cross-view retrieval tasks. Most cross-view hashing methods are developed by assuming that data from different views are well paired, e.g., text-image pairs. In real-world applications, however, this fully-paired multiview setting may not be practical. The more practical yet challenging semi-paired cross-view retrieval problem, where pairwise correspondences are only partially provided, has less been studied. In this paper, we propose an unsupervised hashing method for semi-paired cross-view retrieval, dubbed semi-paired discrete hashing (SPDH). In specific, SPDH explores the underlying structure of the constructed common latent subspace, where both paired and unpaired samples are well aligned. To effectively preserve the similarities of semi-paired data in the latent subspace, we construct the cross-view similarity graph with the help of anchor data pairs. SPDH jointly learns the latent features and hash codes with a factorization-based coding scheme. For the formulated objective function, we devise an efficient alternating optimization algorithm, where the key binary code learning problem is solved in a bit-by-bit manner with each bit generated with a closed-form solution. The proposed method is extensively evaluated on four benchmark datasets with both fully-paired and semi-paired settings and the results demonstrate the superiority of SPDH over several other state-of-the-art methods in term of both accuracy and scalability.

For finding the appropriate number of clusters in short text clustering, models based on Dirichlet Multinomial Mixture (DMM) require the maximum possible cluster number before inferring the real number of clusters. However, it is difficult to choose a proper number as we do not know the true number of clusters in short texts beforehand. The cluster distribution in DMM based on Dirichlet process as prior goes down exponentially as the number of clusters increases. Therefore, we propose a novel model based on Pitman-Yor Process to capture the power-law phenomenon of the cluster distribution in the paper. Specifically, each text chooses one of the active clusters or a new cluster with probabilities derived from the Pitman-Yor Process Mixture model (PYPM). Discriminative words and nondiscriminative words are identified automatically to help enhance text clustering. Parameters are estimated efficiently by collapsed Gibbs sampling and experimental results show PYPM is robust and effective comparing with the state-of-the-art models.

Single sample per person (SSPP) recognition is one of the most challenging problems in face recognition (FR) due to the lack of information to predict the variations in the query sample. To address this problem, we propose in this paper a novel face recognition algorithm based on a robust collaborative representation (CR) and probabilisticgraph model, which is called Collaborative Probabilistic Labels (CPL). First, by utilizing label propagation, we construct probabilistic labels for the samples in the generic training set corresponding to those in the gallery set, thus the discriminative information of the unlabeled data can be effectively explored in our method. Then, the adaptive variation type for a given test sample is automatically estimated. Finally, we propose a novel reconstruction-based classifier for the test sample with its corresponding adaptive dictionary and probabilistic labels. The proposed probabilistic graph based model is adaptively robust to various variations in face images, including illumination, expression, occlusion, pose, etc., and is able to reduce required training images to one sample per class. Experimental results on five widely used face databases are presented to demonstrate the efficacy of the proposed approach. Constructed probabilistic graph propagates discrimination from generic to gallery.The adaptive variation type for a given sample can be automatically estimated.CPL incorporates a novel probabilistic label reconstruction based.

As one of the most popular research topics, sparse representation and dictionary learning technique has received an increasing amount of interest in recent years. Sparse representation based classification (SRC) has been shown to be an effective method and produce impressive performance on face recognition. SRC directly used the entire set of training samples as the dictionary for sparse coding. Recent research has shown that learning a dictionary from the training samples instead of using a predefined one can produce state-of-the-art results. However, all of these dictionary learning methods are designed to achieve low classification errors and implicitly assumes that the losses of all misclassification are the same. In many real-world face recognition applications, this assumption may not hold as different misclassifications could lead to different losses. Motivated by this concern, in this paper we propose a cost-sensitive dictionary learning algorithm for SRC, by which the designed dictionary is able to produce cost-sensitive sparse coding, resulting in improved classification performance in such scenarios. Our method considers the cost information during the sparse coding stages. Specifically, we introduce a new "cost" penalizing matrix and enforce the cost-sensitive requirement throughout the learning process. The optimal solution is efficiently obtained following the alternative optimization method. Experimental results demonstrate the effectiveness of the proposed method. A cost-sensitive dictionary learning algorithm for SRC is proposed.Introduce a new "cost" penalizing matrix during the sparse coding stages.Enforce cost-sensitive requirement throughout the learning process.The learned dictionary is able to produce cost-sensitive sparse coding.Our method can achieve a minimum overall recognition loss.

Locality-based feature learning for multi-view data has received intensive attention recently. As a result of only considering single-category local neighbor relationships, most of such the learning methods are difficult to well reveal intrinsic geometric structure information of raw high-dimensional data. To solve the problem, we propose a novel supervised multi-view correlation feature learning algorithm based on multi-category local neighbor relationships, called multi-patch embedding canonical correlation analysis (MPECCA). Our algorithm not only employs multiple local patches of each raw data to better capture the intrinsic geometric structure information, but also makes intraclass correlation features as close as possible by minimizing intraclass scatter of each view. Extensive experimental results on several real-world image datasets have demonstrated the effectiveness of our algorithm.

Hashing techniques have been widely applied in the large-scale cross-view retrieval tasks due to the significant advantage of hash codes in computation and storage efficiency. Most existing cross-view hashing methods can only handle fully-paired scenarios, where all samples from different views are paired. However, such full pairwise correspondences may not be available in practical applications. In this paper, we propose a novel hashing method, named semi-paired hashing (SPH), to deal with a more challenging cross-view retrieval task, where only partial pairwise correspondences are provided in advance. Specifically, SPH aims to preserve within-view similarity and cross-view correlation among multi-view data. Similarity structure within each view is obtained via anchor graph. As limited samples are paired, correlation between unpaired samples is exploited via a simple yet effective approach, which estimates cross-view correlation by partial cross-view pairwise information and within-view similarity structure. Besides, we further incorporate two regression terms between original features and target binary codes to reduce the quantization loss. An efficient iterative algorithm is presented to simultaneously solve hash functions and binary codes. Extensive experiments on two benchmark datasets demonstrate the superiority of SPH over the state-of-the-art methods, especially in the semi-paired scenarios.

Face hallucination (FH) is an effective technique for super-resolving low-resolution (LR) face images. In real-world applications, a face image usually has multiple distinct low resolutions. Most existing FH methods can not effectively deal with multiple LR views simultaneously. To solve this issue, we present a multi-set partial least squares (MPLS) approach and its kernel extension for jointly learning the nonlinear consistency of multi-resolution facial features. With nonlinear MPLS, we present a novel simultaneous super-resolution coherent facial feature method for the face images with multiple LRs, which has capacity of jointly learning the nonlinear relationships between multiple facial resolutions. Experimental results demonstrate the effectiveness and robustness of our proposed FH method.

Collaborative representation-based classification (CRC) is a famous representation-based classification method in pattern recognition. Recently, many variants of CRC have been designed for many classification tasks with the good classification performance. However, most of them ignore the inter-class pattern discrimination among the class-specific representations, which is very critical for strengthening the pattern discrimination of collaborative representation (CR). In this article, we propose a novel CR approach for image classification, called weighted discriminative collaborative competitive representation (WDCCR). The proposed WDCCR designs the discriminative and competitive collaborative representation among all the classes by fully considering the class information. On the one hand, we incorporate two discriminative constraints into the unified WDCCR model. Both constraints are the competitive class-specific representation residuals and the pairs of class-specific representations for each query sample. On the other hand, the constraint of the weighted categorical representation coefficients is introduced into the proposed model for further enhancing the power of discriminative and competitive representation. In the weighted constraint, we assume that the different classes of each query sample should have less contribution to the representation with the small representation coefficients, and then two types of weight factors are designed to constrain the representation coefficients. Furthermore, the robust WDCCR (R-WDCCR) is proposed with l1-norm representation fidelity for recognizing noisy images. Extensive experiments on six image data sets demonstrate the effective and robust superiorities of the proposed WDCCR and R-WDCCR over the related state-of-the-art representation-based classification methods.

In pattern recognition, the representation-based classification (RBC) has attracted much attention recently. As a representative one of RBC, collaborative representation-based classification (CRC) and its variants have achieved promising classification performance in many visual classification tasks. However, most of the CRC methods cannot directly consider the class discrimination information of data that is very important for classification. To fully use the class discrimination information, we propose a novel discriminative group collaborative competitive representation-based classification method (DGCCR) in this paper. In the designed DGCCR model, the discriminative competitive relationships of classes, the discriminative decorrelations among classes and the weighted class-specific group constraints are simultaneously taken into account for strengthening the power of pattern discrimination. Experiments on three visual classification data sets demonstrate that the proposed DGCCR out-performs state-of-the-art RBC methods.

Face super-resolution (FSR) is an effective way to solve low-resolution (LR) problems in face analysis. But, most FSR methods only consider that LR face images have a single resolution, which is usually not consistent with practical situations due to the existence of multiple resolutions. To date, simultaneously learning the mappings from multiple LRs to high resolution (HR) has not been given proper attention. To solve this issue, we first propose a multi-set partial least squares (MPLS) approach to jointly deal with multi-set random variables via a recursive optimization. With MPLS, we then present a novel FSR method called MPLS-FH to simultaneously learn multiple resolution-specific mappings for various LR views from the same source. Concretely, MPLS-FH first divides multi-resolution face images into many patches. Then, it jointly learns the latent coherent features of principal-component embeddings of multi-resolution patches. Last, it super-resolves the input LR face by cross-resolution neighborhood search. Experimental results demonstrate the effectiveness of the proposed method in terms of quantitative and qualitative evaluations.

Lexical simplification (LS) aims to replace complex words in a given sentence with their simpler alternatives of equivalent meaning. Recently unsupervised lexical simplification approaches only rely on the complex word itself regardless of the given sentence to generate candidate substitutions, which will inevitably produce a large number of spurious candidates. We present a simple LS approach that makes use of the Bidirectional Encoder Representations from Transformers (BERT) which can consider both the given sentence and the complex word during generating candidate substitutions for the complex word. Specifically, we mask the complex word of the original sentence for feeding into the BERT to predict the masked token. The predicted results will be used as candidate substitutions. Despite being entirely unsupervised, experimental results show that our approach obtains obvious improvement compared with these baselines leveraging linguistic databases and parallel corpus, outperforming the state-of-the-art by more than 12 Accuracy points on three well-known benchmarks.

Autonomous vehicles and robot are more and more popular, and it is important to sense the environment for autonomous vehicles and robot. LiDAR is a very powerful device, it can capture the distance and contour of object precisely and quickly. And the scanning range is very wide. we use 3D LiDAR and neural network to detection the environment. We get the data packet from LiDAR, and decode the data packet to 3D point cloud coordinate. Then extract the drivable road by neural network and mark the obstacle.

Canonical correlation analysis (CCA) is a widely used linear unsupervised subspace learning method. However, standard CCA works with vectorized representation of image matrix, which loses the spatial structure information of image data. In addition, a real-world observation often simultaneously belongs to multiple distinct classes with different degrees of membership, while conventional CCA methods can not deal with this situation. Inspired by aforementioned issues, we in this paper propose a fuzzy bilinear canonical correlation projection (FBCCP) approach. FBCCP not only considers two-dimensional spatial structure of images, but also membership degree of practical observation belonging to different classes at the same time. Experimental results on visual recognition show that the proposed FBCCP approach is more effective than related feature learning approaches.

The random subspace method, known as the pillar of random forests, is good at making precise and robust predictions. However, there is not a straightforward way yet to combine it with deep learning. In this paper, we therefore propose Neural Random Subspace (NRS), a novel deep learning based random subspace method. In contrast to previous forest methods, NRS enjoys the benefits of end-to-end, data-driven representation learning, as well as pervasive support from deep learning software and hardware platforms, hence achieving faster inference speed and higher accuracy. Furthermore, as a non-linear component to be encoded into Convolutional Neural Networks (CNNs), NRS learns non-linear feature representations in CNNs more efficiently than previous higher-order pooling methods, producing good results with negligible increase in parameters, floating point operations (FLOPs) and real running time. Compared with random subspaces, random forests and gradient boosting decision trees (GBDTs), NRS achieves superior performance on 35 machine learning datasets. Moreover, on both 2D image and 3D point cloud recognition tasks, integration of NRS with CNN architectures achieves consistent improvements with minor extra cost. Code is available at https://github.com/CupidJay/NRS_pytorch., 38 pages

Canonical correlation analysis (CCA) is a powerful tool for analyzing multi-dimensional paired data. However, when facing semi-supervised multi-modal data (Also called multi-view Hou et al. (Pattern Recog 43(3):720–730, 2010) or multi-represented Kailing et al. (Clustering multi-represented objects with noise. In: Proceedings of the eighth Pacific-Asia conference on knowledge discovery and data mining (PAKDD). Sydney, Australia, pp 394–403) data. For convenience, we will uniformly call them multi-modal data hereafter.) which widely exist in real-world applications, CCA usually performs poorly due to ignoring useful supervised information. Meanwhile, due to the limited labeled training samples in the semi-supervised scenario, supervised extensions of CCA suffer from overfitting. Several semi-supervised extensions of CCA have been proposed recently. Nevertheless, they either just utilize the global structural information captured from the unlabeled data, or propagate label information by discovering the affinities just between the labeled and unlabeled data points in advance. In this paper, we propose a robust multi-modal semi-supervised feature extraction and fusion framework, termed as dual structural consistency based multi-modal correlation propagation projections (SCMCPP). SCMCPP guarantees the consistency between representation structure and hypotaxis structure in each modality and ensures the consistency of hypotaxis structure between two different modalities. By iteratively propagating labels and learning affinities, discriminative information of both given labels and estimated labels is utilized to improve the affinity construction and infer the remaining unknown labels. Moreover, probabilistic within-class scatter matrices in each modality and probabilistic correlation matrix between two modalities are constructed to enhance the discriminative power of features. Extensive experiments on several benchmark face databases demonstrate the effectiveness of our approach.

Multiset canonical correlation analysis (MCCA) is a powerful technique for multi-view joint dimensionality reduction by maximizing linear correlations among the projections. However, most existing MCCA-related methods fail to discover the intrinsic discriminating structure among data spaces and the correspondence between multiple views. In order to address these problems, we incorporate the collaborative representation structure of data points in each view. Then we construct a view-consistent collaborative multiset correlation projection (C2MCP) framework, in which the structures among different views are guaranteed to be consistent and preserved in low-dimensional subspaces. Also, by taking within-class and between-class collaborative reconstruction into account to improve discriminative power for the supervised scenario, we then propose a novel algorithm, called view-consistent collaborative discriminative multiset correlation projection (C2DMCP), to explicitly consider both between-set cumulative correlations and discriminative structure in multiple representation data. The feasibility and effectiveness of the proposed method has been verified on three benchmark databases, i.e., ETH-80, AR and Extended Yale B, with promising results.

Existing kernel-based correlation analysis methods mainly adopt a single kernel in each view. However, only a single kernel is usually insufficient to characterize nonlinear distribution information of a view. To solve the problem, we transform each original feature vector into a 2-dimensional feature matrix by means of kernel alignment, and then propose a novel kernel-aligned multi-view canonical correlation analysis (KAMCCA) method on the basis of the feature matrices. Our proposed method can simultaneously employ multiple kernels to better capture the nonlinear distribution information of each view, so that correlation features learned by KAMCCA can have well discriminating power in real-world image recognition. Extensive experiments are designed on five real-world image datasets, including NIR face images, thermal face images, visible face images, handwritten digit images, and object images. Promising experimental results on the datasets have manifested the effectiveness of our proposed method.