Representation Features Research Articles

Learning representations for image-based profiling of perturbations

Measuring the phenotypic effect of treatments on cells through imaging assays is an efficient and powerful way of studying cell biology, and requires computational methods for transforming images into quantitative data. Here, we present an improved strategy for learning representations of treatment effects from high-throughput imaging, following a causal interpretation. We use weakly supervised learning for modeling associations between images and treatments, and show that it encodes both confounding factors and phenotypic features in the learned representation. To facilitate their separation, we constructed a large training dataset with images from five different studies to maximize experimental diversity, following insights from our causal analysis. Training a model with this dataset successfully improves downstream performance, and produces a reusable convolutional network for image-based profiling, which we call Cell Painting CNN. We evaluated our strategy on three publicly available Cell Painting datasets, and observed that the Cell Painting CNN improves performance in downstream analysis up to 30% with respect to classical features, while also being more computationally efficient.

Репрезентация творчества А. Блока в видеопоэзии

The aim of the research is to identify the most important features in representation and reception of A. Blok’s poetic texts in the genre of video poetry presented at the international festival ‘Videostikhiya’. The paper examines the specifics of video poetry as a new sphere in art associated with the search for both modern forms of poetic creativity and ways of its interpretation and representation; analyzes video clips based on A. Blok’s poems. Scholars’ attention is focused on the complexity of audio visualization of the works by the symbolist poet, who perceived life in its transcendent essence. The scientific novelty of the research lies in the fact that for the first time, the results of analysing modern works in the genre of video poetry based on Blok’s texts are presented; approaches to the study of the visual sphere of A. Blok’s poems are expanded. It is proved that the representation of A. Blok’s creative work in video poetry has its own specifics: the authors can modernize the Blok text, deepening the perception of the poem through metaphorical and symbolic video sequences, or recreate the context, immersing the viewer in the atmosphere of the Silver Age; intertextual inclusions also help each of the video artists to present in the media space a new reading of A. Blok’s poetic statement about the world and about people.

Universal representation learning for multivariate time series using the instance-level and cluster-level supervised contrastive learning

AbstractThe multivariate time series classification (MTSC) task aims to predict a class label for a given time series. Recently, modern deep learning-based approaches have achieved promising performance over traditional methods for MTSC tasks. The success of these approaches relies on access to the massive amount of labeled data (i.e., annotating or assigning tags to each sample that shows its corresponding category). However, obtaining a massive amount of labeled data is usually very time-consuming and expensive in many real-world applications such as medicine, because it requires domain experts’ knowledge to annotate data. Insufficient labeled data prevents these models from learning discriminative features, resulting in poor margins that reduce generalization performance. To address this challenge, we propose a novel approach: supervised contrastive learning for time series classification (SupCon-TSC). This approach improves the classification performance by learning the discriminative low-dimensional representations of multivariate time series, and its end-to-end structure allows for interpretable outcomes. It is based on supervised contrastive (SupCon) loss to learn the inherent structure of multivariate time series. First, two separate augmentation families, including strong and weak augmentation methods, are utilized to generate augmented data for the source and target networks, respectively. Second, we propose the instance-level, and cluster-level SupCon learning approaches to capture contextual information to learn the discriminative and universal representation for multivariate time series datasets. In the instance-level SupCon learning approach, for each given anchor instance that comes from the source network, the low-variance output encodings from the target network are sampled as positive and negative instances based on their labels. However, the cluster-level approach is performed between each instance and cluster centers among batches, as opposed to the instance-level approach. The cluster-level SupCon loss attempts to maximize the similarities between each instance and cluster centers among batches. We tested this novel approach on two small cardiopulmonary exercise testing (CPET) datasets and the real-world UEA Multivariate time series archive. The results of the SupCon-TSC model on CPET datasets indicate its capability to learn more discriminative features than existing approaches in situations where the size of the dataset is small. Moreover, the results on the UEA archive show that training a classifier on top of the universal representation features learned by our proposed method outperforms the state-of-the-art approaches.

DSISA: A New Neural Machine Translation Combining Dependency Weight and Neighbors

Most of the previous neural machine translations (NMT) rely on parallel corpus. Integrating explicitly prior syntactic structure information can improve the neural machine translation. In this article, we propose a Syntax Induced Self-Attention (SISA) which explores the influence of dependence relation between words through the attention mechanism and fine-tunes the attention allocation of the sentence through the obtained dependency weight. We present a new model, Double Syntax Induced Self-Attention (DSISA), which fuses the features extracted by SISA and a compact convolution neural network (CNN). SISA can alleviate long dependency in sentence, while CNN captures the limited context based on neighbors. DSISA utilizes two different neural networks to extract different features for richer semantic representation and replaces the first layer of Transformer encoder. DSISA not only makes use of the global feature of tokens in sentences but also the local feature formed with adjacent tokens. Finally, we perform simulation experiments that verify the performance of the new model on standard corpora.

Simplified Multi-head Mechanism for Few-Shot Remote Sensing Image Classification

The study of few-shot remote sensing image classification has received significant attention. Although meta-learning-based algorithms have been the primary focus of recent examination, feature fusion methods stress feature extraction and representation. Nonetheless, current feature fusion methods, like the multi-head mechanism, are restricted by their complicated network structure and challenging training process. This manuscript presents a simplified multi-head mechanism for obtaining multiple feature representations from a single sample. Furthermore, we perform specific fundamental transformations on remote-sensing images to obtain more suitable features for information representation. Specifically, we reduce multiple feature extractors of the multi-head mechanism to a single one and add an image transformation module before the feature extractor. After transforming the image, the features are extracted resulting in multiple features for each sample. The feature fusion stage is integrated with the classification prediction stage, and multiple linear classifiers are combined for multi-decision fusion to complete feature fusion and classification. By combining image transformation with feature decision fusion, we compare our results with other methods through validation tests and demonstrate that our algorithm simplifies the multi-head mechanism while maintaining or improving classification performance.

High-level aftereffects reveal the role of statistical features in visual shape encoding

Visual shape perception is central to many everyday tasks, from object recognition to grasping and handling tools.1,2,3,4,5,6,7,8,9,10 Yet how shape is encoded in the visual system remains poorly understood. Here, we probed shape representations using visual aftereffects-perceptual distortions that occur following extended exposure to a stimulus.11,12,13,14,15,16,17 Such effects are thought to be caused by adaptation in neural populations that encode both simple, low-level stimulus characteristics17,18,19,20 and more abstract, high-level object features.21,22,23 To tease these two contributions apart, we used machine-learning methods to synthesize novel shapes in a multidimensional shape space, derived from a large database of natural shapes.24 Stimuli were carefully selected such that low-level and high-level adaptation models made distinct predictions about the shapes that observers would perceive following adaptation. We found that adaptation along vector trajectories in the high-level shape space predicted shape aftereffects better than simple low-level processes. Our findings reveal the central role of high-level statistical features in the visual representation of shape. The findings also hint that human vision is attuned to the distribution of shapes experienced in the natural environment.

Deciphering spatial domains from spatially resolved transcriptomics with Siamese graph autoencoder.

Cell clustering is a pivotal aspect of spatial transcriptomics (ST) data analysis as it forms the foundation for subsequent data mining. Recent advances in spatial domain identification have leveraged graph neural network (GNN) approaches in conjunction with spatial transcriptomics data. However, such GNN-based methods suffer from representation collapse, wherein all spatial spots are projected onto a singular representation. Consequently, the discriminative capability of individual representation feature is limited, leading to suboptimal clustering performance. To address this issue, we proposed SGAE, a novel framework for spatial domain identification, incorporating the power of the Siamese graph autoencoder. SGAE mitigates the information correlation at both sample and feature levels, thus improving the representation discrimination. We adapted this framework to ST analysis by constructing a graph based on both gene expression and spatial information. SGAE outperformed alternative methods by its effectiveness in capturing spatial patterns and generating high-quality clusters, as evaluated by the Adjusted Rand Index, Normalized Mutual Information, and Fowlkes-Mallows Index. Moreover, the clustering results derived from SGAE can be further utilized in the identification of 3-dimensional (3D) Drosophila embryonic structure with enhanced accuracy. Benchmarking results from various ST datasets generated by diverse platforms demonstrate compelling evidence for the effectiveness of SGAE against other ST clustering methods. Specifically, SGAE exhibits potential for extension and application on multislice 3D reconstruction and tissue structure investigation. The source code and a collection of spatial clustering results can be accessed at https://github.com/STOmics/SGAE/.

Multi-task Adaptive Resolution Network for Lymph Node Metastasis Diagnosis from Whole Slide Images of Colorectal Cancer.

Automated detection of lymph node metastasis (LNM) holds great potential to alleviate the workload of doctors and reduce misinterpretations. Despite the practical successes achieved, effectively addressing the highly complex and heterogeneous tumor microenvironment remains an open and challenging problem, especially when tumor subtypes intermingle and are difficult to delineate. In this paper, we propose a multi-task adaptive resolution network, named MAR-Net, for LNM detection and subtyping in complex mixed-type cancers. Specifically, we construct a resolution-aware module to mine heterogeneous diagnostic information, which exploits the multi-scale pyramid information and adaptively combines multi-resolution structured features for comprehensive representation. Additionally, we adopt a multi-task learning approach that simultaneously addresses LNM detection and subtyping, reducing model instability during optimization and improving performance across both tasks. More importantly, to rectify the potential misclassification of tumor subtypes, we elaborately design a hierarchical subtying refinement (HSR) algorithm that leverages a generic segmentation model informed by pathologists' prior knowledge. Evaluations have been conducted on three private and one public cancer datasets (554 WSIs, 4.8 million patches). Our experimental results demonstrate that the proposed method consistently achieves superior performance compared to the state-of-the-art methods, achieving 0.5% to 3.2% higher AUC in LNM detection and 3.8% to 4.4% higher AUC in LNM subtyping.

Dynamic Graph Regularized Broad Learning With Marginal Fisher Representation for Noisy Data Classification.

Broad learning system (BLS) is an effective neural network requiring no deep architecture, however it is somehow fragile to noisy data. The previous robust broad models directly map features from the raw data, which inevitably learn useless or even harmful features for data representation when the inputs are corrupted by noise and outliers. To address this concern, a discriminative and robust network named as dynamic graph regularized broad learning (DGBL) with marginal fisher representation is proposed for noisy data classification. Different from the previous works, DGBL eliminates the effect of noise before the random feature mapping by the proposed robust and dynamic marginal fisher analysis (RDMFA) algorithm. The RDMFA is able to extract more robust and informative representations for classification from the latent clean data space with dynamically generated graphs. Furthermore, the dynamic graphs learned from RDMFA are incorporated as regularization terms into the objective of DGBL to enhance the discrimination capacity of the proposed network. Extensive quantitative and qualitative experiments conducted on numerous benchmark datasets demonstrate the superiority of the proposed model compared to several state-of-the-art methods.

Low-Shot Unsupervised Visual Anomaly Detection via Sparse Feature Representation.

Visual anomaly detection is an essential component in modern industrial manufacturing. Existing studies using notions of pairwise similarity distance between a test feature and nominal features have achieved great breakthroughs. However, the absolute similarity distance lacks certain generalizations, making it challenging to extend the comparison beyond the available samples. This limitation could potentially hamper anomaly detection performance in scenarios with limited samples. This article presents a novel sparse feature representation anomaly detection (SFRAD) framework, which formulates the anomaly detection as a sparse feature representation problem; and notably proposes an anomaly score by orthogonal matching pursuit (ASOMP) as a novel detection metric. Specifically, SFRAD calculates the Gaussian kernel distance between the test feature and its sparse representation in the nominal feature space for anomaly detection. Here, the orthogonal matching pursuit (OMP) algorithm is adopted to achieve the sparse feature representation. Moreover, to construct a low-redundancy memory bank storing the basis features for sparse representation, a novel basis feature sampling (BFS) algorithm is proposed by considering both the maximum coverage and the optimum feature representation simultaneously. As a result, SFRAD incorporates both the advantages of absolute similarity and linear representation; and this enhances the generalization in low-shot scenarios. Extensive experiments on the MVTec anomaly detection (MVTec AD), Kolektor surface-defect dataset (KolektorSDD), Kolektor surface-defect dataset 2 (KolektorSDD2), MVTec logical constraints anomaly detection (MVTec LOCO AD), Visual anomaly (VISA), Modified national institute of standards and technology (MNIST), and CIFAR-10 datasets demonstrate that our proposed SFRAD outperforms the previous methods and achieves state-of-the-art unsupervised anomaly detection performance. Notably, significantly improved outcomes and results have also been achieved on low-shot anomaly detection. Code is available at https://github.com/fanghuisky/SFRAD.

Infrared Visible Image Registration Based on Gabor Representation Feature Descriptor

Infrared Visible Image Registration Based on Gabor Representation Feature Descriptor

National-Cultural Specificity of Biblical Idioms

The article investigates the representation features of idioms of biblical origin on the material of various languages. Idioms of biblical origin are regarded as linguocultural and nationally specific units, reflecting the peculiarities of mentality and religious worldview of a particular linguistic culture. The article reveals a number of extralinguistic factors that determine both the universal and specific features of idioms of biblical origin representation in various languages. The factors contributing to the universality of idioms of biblical origin include: 1) the commonality of Christian culture; 2) the influence of the Greek and Latin languages on subsequent translations of the Bible used nowadays. The followings are singled out as extralinguistic factors that determine the specificity of idioms of biblical origin in various languages: 1) the emergence time and dissemination scope of biblical translations; 2) confessional affiliation of a certain ethnic group; 3) peculiarities of national perception of the world and cultural values of a particular linguistic culture; 4) the socio-political system of a particular era or country; 5) mass-media factors. The national-cultural specificity of idioms of biblical origin is manifested in their non-equivalence in a particular language, the quantitative divergence of idioms of biblical origin included into a particular language from a single source, differences in the semantic capacity and meaning of idioms of biblical origin used in each of the languages, the emphasis on certain images and cognitive features activated in idioms of biblical origin, in existance of a set of borrowed idioms of biblical origin from different languages. It should be noted that even universal idioms of biblical origin functioning in different languages can be nationally specific due to the differences in the activated semantic features and figurative components of idioms.

Investigating the impact of representation features on decision model comprehension

ABSTRACT Decision models play an important role in various areas of information systems research, including system analysis and design, compliance management, and various application domains. Decision models must be effectively presented so that analysts can assure their correctness and completeness. So far, empirical research on the cognitive effectiveness of decision models has provided partially inconclusive results. Our paper provides novel insights into the drivers of decision model comprehension by moving from a classification based on representation archetypes to granular representation features. Using an experimental research design, we discover that the decision model type must be assessed in conjunction with other representational factors, such as representation structure (expanded vs. frugal) and representation design (monochromatic vs. colours). In this way, we extend prior arguments of cognitive fit theory by demonstrating that colour can be used to compensate for a misfit of decision model and task, and structural features can further increase model comprehension. We further studied the root causes of the observed effects by using eye-tracking. Our findings have implications for both cognitive information systems research and practice, as they can be used to guide decision model users and tool vendors.

Deep Fuzzy Variable C-Means Clustering Incorporated With Curriculum Learning

End-to-end deep clustering method utilizes deep neural networks to jointly learn representation features and clustering assignments. Although many k-means-friendly deep clustering models have been explored, the existing division-based methods tend to directly implement clustering with a specific number of clusters, which suffers from poor performance resulted from indistinguishable clusters, and contributes to bad local optimum. At the same time, the representation learning of fuzzy <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$c$</tex-math></inline-formula> -means clustering in the feature space still needs more research. In this paper, a Deep Fuzzy Curriculum Clustering (DFC <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> ) method with the learning strategy of clustering from easy to complex automatically is proposed to tackle above issues. Firstly, considering the soft flexible allocation of fuzzy <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$c$</tex-math></inline-formula> -means and the preservation of local structure of original data, the fuzzy clustering loss and the autoencoder's reconstruction loss are constructed to learn the embedded features and clustering centers simultaneously. Second, curriculum loss is introduced into the constraint to make clusters successively merge in line with implementing clustering from easy to complex, and realize the bottom-up deep aggregative clustering automatically. In addition, novel curriculum information is proposed as constraint to guide the merging of clusters belonging to the same class. Experimental results on four real-world datasets show the superiority of the proposal.

Transfer Learning for Indoor Localization Algorithm Based on Deep Domain Adaptation

In application, training data and test data collected via indoor positioning algorithms usually do not come from the same ideal conditions. Changes in various environmental conditions and signal drift can cause different probability distributions between the data sets. Existing positioning algorithms cannot guarantee stable accuracy when facing these issues, resulting in dramatic reduction and the infeasibility of the positioning accuracy of indoor location algorithms. Considering these restrictions, domain adaptation technology in transfer learning has proven to be a promising solution in past research in terms of solving the inconsistent probability distribution problems. However, most localization algorithms based on transfer learning do not perform well because they only learn a shallow representation feature, which can only slightly reduce the domain discrepancy. Based on the deep network and its strong feature extraction ability, it can learn more transferable features for domain adaptation and achieve better domain adaptation effects. A Deep Joint Mean Distribution Adaptation Network (DJMDAN) is proposed to align the global domain and relevant subdomain distributions of activations in multiple domain-specific layers across domains to achieve domain adaptation. The test results demonstrate that the performance of the proposed method outperforms the comparison algorithm in indoor positioning applications.

Mutual Boost Network for attributed graph clustering

Attributed graph clustering is an essential research topic on real-world data. However, the heterogeneous gap between node and structure features limits the existing approaches for discriminative representation. Besides, how to ensure the consistency of clustering assignments produced by the node and structure features is still a challenging problem, which always causes inferior clustering performance. To overcome above problems, a dual-channel network for attributed graph clustering named Mutual Boost Network (MBN) is proposed, which consists of auto-encoder and graph auto-encoder that can interact and learn from each other to achieve mutual boost of clustering performance. First, a novel representation enhancement module is proposed to disseminate the heterogeneous information from both node and structure features for learning comprehensive representation. Then, a consistency constraint by contrasting clustering assignments is devised to provide a mutual guide to make them tend to be consistent. Finally, the procedures of representation learning and clustering assignment are simultaneously optimized under a unified framework in a self-supervised manner. Experimental results on six graph collections show that the MBN performs better than the SOTA approaches by a statistically significant margin. The MBN code can be found at: https://github.com/Xiaoqiang-Yan/MBN.

A Lightweight Multimodal Footprint Recognition Network Based on Progressive Multi-Granularity Feature Fusion

The main differences in images of footprints are the proportion of the parts of foot and the distribution of pressure, which can be considered as fine-grained image classification. Moreover, the deviation of human body weight and muscle strength increases the difficulty of identifying the left and right feet. While using a fine-grained image classification network to solve the footprint image classification problem is certainly a feasible approach, the number of parameters in a fine-grained image classification network is generally large, and therefore we would like to build a lightweight classification network that is suitable for several small footprint datasets. In this paper, a multimodal footprint recognition algorithm based on progressive multi-granularity feature fusion is proposed. First, the shallow dense connection network is used to extract features. The feature extraction ability of the model is improved with the help of channel splicing and feature multiplexing. Second, to learn footprint images of different granularities, the progressive training strategy and puzzle scrambler are applied to the model. Finally, factorized bilinear coding can aggregate local features to obtain more discriminative global representation features. Experiments show that our network achieves comparable classification accuracy to some fine-grained image classification models (PMG, MSEC) on the complete pressure footprint dataset, but the number of parameters in our network is greatly reduced. Meanwhile, our network also achieves good classification results on several other footprint datasets, which demonstrates the effectiveness of our network. At the same time, an ablation experiment was carried out to verify the effectiveness of the progressive strategy and the factorized bilinear coding.

ASENN: attention-based selective embedding neural networks for road distress prediction

This study proposes an innovative neural network framework, ASENN (Attention-based Selective Embedding Neural Network), for the prediction of pavement deterioration. Considering the complexity and uncertainty associated with the pavement deterioration process, two fundamental frameworks, SEL (Selective Embedding Layer) and MDAL (Multi-Dropout Attention Layer), are combined to enhance feature abstraction and prediction accuracy. This approach is significant while analyzing the pavement deterioration process due to the high variability of the contributing deterioration factors. These factors, represented as tabular data, undergo filtering, embedding, and fusion stages in the SEL, to extract crucial features for an effective representation of pavement deterioration. Further, multiple attention-weighted combinations of raw data are obtained through the MDAL. Several SELs and MDALs were combined as basic cells and layered to form an ASENN. The experimental results demonstrate that the proposed model outperforms existing tabular models on four road distress parameter datasets corresponding to cracking, deflection, international roughness index, and rutting. The optimal number of cells was determined using different ablation settings. The results also show that the feature learning capabilities of the ASENN model improved as the number of cells increased; however, owing to the limited combination space of feature fields, extreme depths were not preferred. Furthermore, the ablation investigation demonstrated that MDAL can improve performance, particularly on the cracking dataset. Notably, compared with mainstream transformer models, ASENN requires significantly less storage and achieves faster execution speed.

Deep reinforcement learning-based resilience enhancement strategy of unmanned weapon system-of-systems under inevitable interferences

Unmanned weapon system-of-systems (UWSoS) is a collection of unmanned weapon systems providing multiple interdependent capabilities to support the mission completion. The soaring number of interconnected systems makes UWSoS vulnerable in the face of inevitable uncertain interferences. Thus, devising an effective resilience enhancement strategy is critical to handling inevitable disruption events. However, preventive and protective methods are not all-inclusive because of the high uncertainty operation environment and the lack of autonomy. Hence, we studied the resilience enhancement problem of UWSoS from the recovery perspective based on deep reinforcement learning (DRL). First, a DRL-based resilience enhancement strategy framework is proposed, combining the graph convolution network and proximal policy optimization algorithm to extract the entities’ representation features and autonomously learn the resilience enhancement strategy to handle various interferences scenarios better. Subsequently, a collaboration action resilience contribution index-guided proximal policy optimization algorithm is proposed to improve training efficiency. Finally, extensive simulation experiments and comparisons with five similar algorithms demonstrate the effectiveness, adaptability, and superiority of the proposed strategy. This work could provide valuable scheduling schemes for decision-makers to guide the reliable operation of UWSoSs.

Efficient Facial Expression Recognition With Representation Reinforcement Network and Transfer Self-Training for Human–Machine Interaction

The emotion perception capability could endow machines with deeper feedback evaluation of human–machine interaction (HMI), which makes the experience of HMI more harmonious and natural. Currently, how to obtain the robust surface and semantic facial representation features has not been fully explored, moreover, the parameters and computation of conventional convolutional neural network models, which are adopted in facial expression recognition (FER) are too heavy, it is not conducive to the deployment of the HMI system. Therefore, this article proposes a representation reinforcement network (RRN) and transfer self-training (TST) based efficient FER method. Our designed RRN is mainly composed of two modules, i.e., surface representation reinforcement (SurRR) module and semantic representation reinforcement (SemaRR) module. SurRR module fuses the facial features in different dimensions and highlights key feature messaging nodes in feature maps. SemaRR module establishes the deep semantic representation of global facial regions on spatial and channel dimensions. Therefore, RRN has stronger feature extraction capability when the parameters and computational complexity are significantly reduced, and TST is proposed for contributing to the training effect of RRN at the condition of insufficient training data. We tested our method on CK+, RaFD, FERPLUS, and RAF-DB datasets, and the recognition accuracy are 100%, 98.62%, 89.64%, and 88.72%, respectively. Furthermore, the preliminary application experiment, which demonstrates the feasibility of our method in practice during HMI.