Noisy Information Research Articles

Harnessing Representative Spatial-Temporal Information for Video Question Answering

Video question answering, aiming to answer a natural language question related to the given video, has become prevalent in the past few years. Although remarkable improvements have been obtained, it is still exposed to the challenge of insufficient comprehension of video content. To this end, we propose a spatial-temporal representative visual exploitation network for video question answering, which enhances the understanding of the video by merely summarizing representative visual information. In order to explore representative object information, we advance adaptive attention based on uncertainty estimation. At the same time, to capture representative frame-level and clip-level visual information, we structure a much more compact set of representations iteratively in an expectation-maximization manner to deprecate noisy information. Both the quantitative and qualitative results on NExT-QA, TGIF-QA, MSRVTT-QA, and MSVD-QA datasets demonstrate the superiority of our model over several state-of-the-art approaches.

Second Order Extended Ensemble Filter for Non-linear Filtering

Whenever the state of a system must be estimated from noisy information, a state estimator is employed to fuse the data with the model to produce an accurate estimate of the state. When the system dynamics and observation models are linear, the Kalman Filter which is optimal, is used. However, in most applications of interest the system dynamics and observations equations are not- linear and suitable extensions of the Kalman Filter have been developed; for example, the Extended Kalman Filter(EKF). The Extended Kalman Filter is based on linearization by the Taylor series expansion about the mean of the state. This filtering process is however computationally expensive especially in high dimensional data. The cause for this is the high cost of integrating the equation of evolution of covariances. Due to this complexity in integration, new methods were sought known as the particle filters. It replaces linearization of non-linearities with Monte Carlo methods. The particle filter formed a basis for Ensemble Kalman Filter (EnKF) an extension of Kalman filter to non-linear filtering. The EnKF reduced the computational cost but its innovation process does not capture information sufficiently hence there is need to improve its performance. This study has developed a new filter, Second order Extended Ensemble Filter (SoEEF).We derived it from stochastic state models by expansion of expected values to the second order by use of Taylor series together with Monte Carlo method and Matlab. We used Lorenz 63 system of ordinary equations and differential Matlab to test the performance of the new filter. Then we compared its performance with four other filters like Bootstrap Particle Filter (BPF), First order Kalman Bucy Filter (FoEKBF),Second order Kalman Bucy Filter (SoKBF) and First order Extended Ensemble Filter (FoEEF). SoEEKF performs much better than the other four filters.

Assessing the performance of a hybrid max‐weight traffic signal control algorithm in the presence of noisy queue information: An evaluation of the environmental impacts

Assessing the performance of a hybrid max‐weight traffic signal control algorithm in the presence of noisy queue information: An evaluation of the environmental impacts

Virtual reality-enabled high-performance emotion estimation with the most significant channel pairs

Human-computer interface (HCI) and electroencephalogram (EEG) signals are widely used in user experience (UX) interface designs to provide immersive interactions with the user. In the context of UX, EEG signals can be used within a metaverse system to assess user engagement, attention, emotional responses, or mental workload. By analyzing EEG signals, system designers can tailor the virtual environment, content, or interactions in real time to optimize UX, improve immersion, and personalize interactions. However, in this case, in addition to the signals' processing cost and classification accuracy, cybersickness in Virtual Reality (VR) systems needs to be resolved. At this point, channel selection methods can perform better for HCI and UX applications by reducing noisy and redundant information in generally unrelated EEG channels. For this purpose, a new method for EEG channel selection based on phase-locking value (PLV) analysis is proposed. We hypothesized that there are interactions between EEG channels in terms of PLV in repeated tasks in different trials of the emotion estimation experiment. Subsequently, frequency-based features were extracted. The features were classified by dividing them into bags using the Multiple-Instance Learning (MIL) variant. This study provides higher classification performance using fewer EEG channels for emotion prediction. The performance rate obtained in binary classification with the Random Forests (RF) algorithm is at a promising level of 99%. The proposed method achieved an accuracy of 99.38% for valence using all channels on the new dataset (VREMO) and 98.13% with channel selection. The benchmark dataset (DEAP) achieved accuracies of 98.16% using all channels and 98.13% with selected channels.

Weakly Supervised Underwater Object Real-time Detection Based on High-resolution Attention Class Activation Mapping and Category Hierarchy

Recently, deep learning-based underwater object detection technology has achieved remarkable success. However, the accuracy and completeness of dataset instance annotation are crucial for its success. The quality of underwater images is low, severe objects clustering, and occlusion, acquiring object's annotations demands substantial time and labor costs, while mis annotation and missed annotation can also degrade model performance and limit their application in practical scenarios. To address this issue, this paper presents a novel weakly supervised underwater object real-time detection method, which is divided into two subtasks: weakly supervised object localization and real-time object detection. In the weakly supervised object localization task, we design a novel category hierarchy structure network that integrates the high-resolution attention-class activation mapping algorithm to obtain high-quality object class activation maps, weaken background interference, and obtain more complete object regions. The parameterized spatial loss module is devised to enable the model to escape from local optimal solutions, thus accurately and efficiently obtaining object pseudo-detection annotation boxes. For the real-time object detection task, the single-stage detector YOLOv7 is selected as the basic detection model, and an object perception loss function is designed based on the class activation map to jointly supervise the training process. A method for filtering noisy pseudo-supervision information is proposed to enhance the pseudo-supervision information involved in training. Ablation experiments and multi-method comparison experiments were conducted on the URPC and RUOD datasets, and the results verify the effectiveness of the proposed strategy, and our model exhibits significant advantages in detection performance and detection efficiency compared to current mainstream and advanced models.

Atom: An Efficient Query Serving System for Embedding-based Knowledge Graph Reasoning with Operator-level Batching

Knowledge graph reasoning (KGR) answers logical queries over a knowledge graph (KG), and embedding-based KGR (EKGR) becomes popular recently, which embeds both queries and KG entities such that the vector embeddings of a query and its answer entities are similar. Compared with traditional KGR methods based on subgraph matching, EKGR produces fewer intermediate results and is more robust to missing and noisy information in the KG. However, existing systems are inefficient for serving online EKGR queries because they can only batch queries of the same type for execution (i.e., query-level batching ) and hence have limited batching opportunities due to the heterogeneity of queries. To serve EKGR queries efficiently, we propose the Atom system with operator-level batching, which decomposes queries into operators and batches operators of the same type from different queries for execution. The insight is that the types of operators are far fewer than the types of queries, and thus different queries typically share common operators, yielding more batching opportunities. To schedule the operators, Atom adopts a hybrid policy, which improves system throughput and avoids starving rare operators. For efficiency, Atom incorporates system optimizations including two-level pipeline, opportunistic submission, pre-allocated memory buffer, and tailored GPU kernels. Experiment results show that compared with existing systems, Atom can improve query throughput by over 20x and reduce query latency by over 5x. Micro experiments suggest that the designs and optimizations are effective in improving system performance.

Auto-focus tracing: Image manipulation detection with artifact graph contrastive

Existing image manipulation detection methods rely on manipulation traces, leading to sub-optimal performance. This is primarily due to their challenges in distinguishing manipulation traces amidst extensive noisy information in the entire image and overlooking artifact relations between manipulated and authentic regions. Toward this end, we propose an Auto-Focus Graph Contrastive Learning, called AFGCL, for image manipulation detection, incorporating two innovative modules: multi-scale view generation (MSVG) and artifact relations modeling (ARM). The MSVG module focuses on generating a pair of contrastive views, each encompassing a different scale region of interest, namely the manipulated region and its surrounding authentic region. Meanwhile, the ARM module emphasizes modeling artifact relations between the pixels of the region of interest to learn a discriminative representation. By minimizing the distance between representations and artifact relations of corresponding views, the AFGCL automatically focuses on the region of interest. This enables a thorough exploration of artifact relations, facilitating accurate manipulation detection. The results show that AFGCL outperforms previous state-of-the-art methods on widely used benchmarks. Code is available at: https://github.com/pwy-cmd/AFGCL.

Hyperspectral Attention Network for Object Tracking

Hyperspectral video provides rich spatial and spectral information, which is crucial for object tracking in complex scenarios. Despite extensive research, existing methods often face an inherent trade-off between rich spectral information and redundant noisy information. This dilemma arises from the efficient utilization of hyperspectral image data channels. To alleviate this problem, this paper introduces a hierarchical spectral attention network for hyperspectral object tracking. We employ a spectral band attention mechanism with adaptive soft threshold to examine the correlations across spectral bands, which integrates the information available in various spectral bands and eliminates redundant information. Moreover, we integrate spectral attention into a hierarchical tracking network to improve the integration of spectral and spatial information. The experimental results on entire public hyperspectral competition dataset WHISPER2020 show the superior performance of our proposed method compared with that of several related methods in visual effects and objective evaluation.

Simultaneous image patch attention and pruning for patch selective transformer

Vision transformer models provide superior performance compared to convolutional neural networks for various computer vision tasks but require increased computational overhead with large datasets. This paper proposes a patch selective vision transformer that effectively selects patches to reduce computational costs while simultaneously extracting global and local self-representative patch information to maintain performance. The inter-patch attention in the transformer encoder emphasizes meaningful features by capturing the inter-patch relationships of features, and dynamic patch pruning is applied to the attentive patches using a learnable soft threshold that measures the maximum multi-head importance scores. The proposed patch attention and pruning method provides constraints to exploit dominant feature maps in conjunction with self-attention, thus avoiding the propagation of noisy or irrelevant information. The proposed patch-selective transformer also helps to address computer vision problems such as scale, background clutter, and partial occlusion, resulting in a lightweight and general-purpose vision transformer suitable for mobile devices.

Deep unsupervised shadow detection with curriculum learning and self-training

Shadow detection is undergoing a rapid and remarkable development along with the wide use of deep neural networks. Benefiting from a large number of training images annotated with strong pixel-level ground-truth masks, current deep shadow detectors have achieved state-of-the-art performance. However, it is expensive and time-consuming to provide the pixel-level ground-truth mask for each training image. Considering that, this paper proposes the first unsupervised deep shadow detection framework, which consists of an initial pseudo label generation (IPG) module, a curriculum learning (CL) module and a self-training (ST) module. The supervision signals used in our learning framework are generated from several existing traditional unsupervised shadow detectors, which usually contain a lot of noisy information. Therefore, each module in our unsupervised framework is dedicated to reduce the adverse influence of noisy information on model training. Specifically, the IPG module combines different traditional unsupervised shadow maps to obtain their complementary shadow information. After obtaining the initial pseudo labels, the CL module and the ST module will be used in conjunction to gradually learn new shadow patterns and update the qualities of pseudo labels simultaneously. Extensive experimental results on various benchmark datasets demonstrate that our deep shadow detector not only outperforms the traditional unsupervised shadow detection methods by a large margin but also achieves comparable results with some recent state-of-the-art fully-supervised deep shadow detection methods.

Cluster-based Graph Collaborative Filtering

Graph Convolution Networks (GCNs) have significantly succeeded in learning user and item representations for recommendation systems. The core of their efficacy is the ability to explicitly exploit the collaborative signals from both the first- and high-order neighboring nodes. However, most existing GCN-based methods overlook the multiple interests of users while performing high-order graph convolution. Thus, the noisy information from unreliable neighbor nodes ( \(e.g.\) , users with dissimilar interests) negatively impacts the representation learning of the target node. Additionally, conducting graph convolution operations without differentiating high-order neighbors suffers the over-smoothing issue when stacking more layers, resulting in performance degradation. In this paper, we aim to capture more valuable information from high-order neighboring nodes while avoiding noise for better representation learning of the target node. To achieve this goal, we propose a novel GCN-based recommendation model, termed Cluster-based Graph Collaborative Filtering (ClusterGCF). This model performs high-order graph convolution on cluster-specific graphs, which are constructed by capturing the multiple interests of users and identifying the common interests among them. Specifically, we design an unsupervised and optimizable soft node clustering approach to classify user and item nodes into multiple clusters. Based on the soft node clustering results and the topology of the user-item interaction graph, we assign the nodes with probabilities for different clusters to construct the cluster-specific graphs. To evaluate the effectiveness of ClusterGCF, we conducted extensive experiments on four publicly available datasets. Experimental results demonstrate that our model can significantly improve recommendation performance.

Multi-attention network with redundant information filtering for multi-horizon forecasting in multivariate time series

Multi-horizon forecasting of multivariate time series has always been a prominent research topic in domains such as finance and transportation. While prediction models that integrate attention mechanisms and Seq2Seq frameworks effectively capture the temporal dependencies between sequences, there might exist distinctions in the contributions of different input variables to the predictions. RNN-based encoders often overlook these distinctions. Hence, we propose a Multi-attention Network with Redundant Information Filtering (MNRIF) for Multi-horizon Forecasting in Multivariate Time Series, comprising the underlying component Two-stage Feature Fusion Network (TSFFN) and the upper-level structure Temporal Attention Network(TAN). TSFFN first designs a variable-level attention to screen the spatial semantics of explanatory variables during multi-step recursive iterations, effectively reducing noisy information, then builds a feature gate fusion module to adaptively weigh the degree of association between prediction targets and spatial semantics as well as target variables, thereby accomplishing secondary filtering of redundant information. TAN is a task-oriented learner that combines the Encoder–Decoder framework with attention mechanism. By means of dual information filtering based on attention mechanism and recurrent gate, it enhances effective mining of long-term temporal dependency features in sequences. Extensive experiments conducted on three real-world benchmark datasets demonstrate MNRIF’s superior performance compared to various state-of-the-art methods of different types.

Implementation and Efficient Analysis of Preprocessing Techniques in Deep Learning for Image Classification

Deep learning models have recently been preferred to perform certain image-processing tasks. Recently, with the increasing radiation, heat, and poor lighting conditions, the raw image samples may contain noisy and ambiguous information. To process these images, the deep learning model requires a large number of data samples to learn the missing information from other clear data samples. This necessitates training the neural network with a huge dataset. The researchers are now attempting to filter and improve such noisy images via preprocessing in order to provide valid and accurate feature information to the neural network layers. However, certain research studies claim that some useful information may be lost when the image is not preprocessed with an appropriate filter or enhancement technique. The MSA (meta-synthesis and analysis) approach is utilized in this work to present the impact of the image processing applications done with and without preprocessing steps. Also, this work summarizes the existing deep learning-based image processing models utilizing or not preprocessing steps in their implementation. This work has also found that 85% of the existing techniques involve a preprocessing step while developing a deep learning model. However, a maximum accuracy of 96.89% is observed on Sine-Net when it is implemented without a preprocessing and the same model gave 96.85% when implemented with preprocessing. This research provides various research insights on the requirement and non-requirement of preprocessing steps in a deep learning-based implementation.

Drug–target interaction prediction through fine-grained selection and bidirectional random walk methodology

The study of drug–target interaction plays an important role in the process of drug development. The subject of DTI forecasting has advanced significantly in the last several years, yielding numerous significant research findings and methodologies. Heterogeneous data sources provide richer information and comprehensive perspectives for drug–target interaction prediction, so many existing methods rely on heterogeneous networks, and graph embedding technology becomes an important technology to extract information from heterogeneous networks. These approaches, however, are less concerned with potential noisy information in heterogeneous networks and more focused on the extent of information extraction in those networks. Based on this, a potential DTI predictive network model called FBRWPC is proposed in this paper. It uses a fine-grained similarity selection program to first integrate similarity on similar networks and then a bidirectional random walk graph embedding learning method with restart to obtain an updated drug target interaction matrix. Through the use of similarity selection and fine-grained selection similarity integration, the framework can effectively filter out the noise present in heterogeneous networks and enhance the model's prediction performance. The experimental findings demonstrate that, even after being split up into four distinct types of data sets, FBRWPC can still retain great prediction performance, a sign of the model's resilience and good generalization.

Leveraged Matrix Completion With Noise.

Completing low-rank matrices from subsampled measurements has received much attention in the past decade. Existing works indicate that O(nrlog2(n)) datums are required to theoretically secure the completion of an n ×n noisy matrix of rank r with high probability, under some quite restrictive assumptions: 1) the underlying matrix must be incoherent and 2) observations follow the uniform distribution. The restrictiveness is partially due to ignoring the roles of the leverage score and the oracle information of each element. In this article, we employ the leverage scores to characterize the importance of each element and significantly relax assumptions to: 1) not any other structure assumptions are imposed on the underlying low-rank matrix and 2) elements being observed are appropriately dependent on their importance via the leverage score. Under these assumptions, instead of uniform sampling, we devise an ununiform/biased sampling procedure that can reveal the "importance" of each observed element. Our proofs are supported by a novel approach that phrases sufficient optimality conditions based on the Golfing scheme, which would be of independent interest to the wider areas. Theoretical findings show that we can provably recover an unknown n×n matrix of rank r from just about O(nrlog2 (n)) entries, even when the observed entries are corrupted with a small amount of noisy information. The empirical results align precisely with our theories.

Attribute reduction with fuzzy divergence-based weighted neighborhood rough sets

Neighborhood rough sets are well-known as an interesting approach for attribute reduction in numerical/continuous data tables. Nevertheless, in most existing neighborhood rough set models, all attributes are assigned the same weights. This may undermine the capacity to select important attributes, especially for high-dimensional datasets. To establish attribute weights, in this study, we will utilize fuzzy divergence to evaluate the distinction between each attribute with the whole attributes in classifying the objects to the decision classes. Then, we construct a new model of fuzzy divergence-based weighted neighborhood rough sets, as well as propose an efficient attribute reduction algorithm. In our method, reducts are considered under the scenario of the α-certainty region, which is introduced as an extension of the positive region. Several related properties will show that attribute reduction based on the α-certainty region can significantly enhance the ability to identify optimal attributes due to reducing the influence of noisy information. To validate the effectiveness of the proposed algorithm, we conduct experiments on 12 benchmark datasets. The results demonstrate that our algorithm not only significantly reduces the number of attributes compared to the original data but also enhances classification accuracy. In comparison to some other state-of-the-art algorithms, the proposed algorithm also outperforms in terms of classification accuracy for almost all of datasets, while also maintaining a highly competitive reduct size and computation time.

Cube is a good form: Hyperspectral band selection via multi-dimensional and high-order structure preserved clustering

As an effective strategy for reducing the noisy and redundant information for hyperspectral imagery (HSI), hyperspectral band selection intends to select a subset of original hyperspectral bands, which boosts the subsequent different tasks. In this paper, we introduce a multi-dimensional high-order structure preserved clustering method for hyperspectral band selection, referred to as MHSPC briefly. By regarding original hyperspectral images as a tensor cube, we apply the tensor CP (CANDECOMP/PARAFAC) decomposition on it to exploit the multi-dimensional structural information as well as generate a low-dimensional latent feature representation. In order to capture the local geometrical structure along the spectral dimension, a graph regularizer is imposed on the new feature representation in the lower dimensional space. In addition, since the low rankness of HSIs is an important global property, we utilize a nuclear norm constraint on the latent feature representation matrix to capture the global data structure information. Different to most of previous clustering based hyperspectral band selection methods which vectorize each band as a vector without considering the 2-D spatial information, the proposed MHSPC can effectively capture the spatial structure as well as the spectral correlation of original hyperspectral cube in both local and global perspectives. An efficient alternatively updating algorithm with theoretical convergence guarantee is designed to solve the resultant optimization problem, and extensive experimental results on four benchmark datasets validate the effectiveness of the proposed MHSPC over other state-of-the-arts.

SA-ConvNeXt: A Hybrid Approach for Flower Image Classification Using Selective Attention Mechanism

In response to the current lack of annotations for flower images and insufficient focus on key image features in traditional fine-grained flower image classification based on deep learning, this study proposes the SA-ConvNeXt flower image classification model. Initially, in the image preprocessing stage, a padding algorithm was used to prevent image deformation and loss of detail caused by scaling. Subsequently, the model was integrated using multi-level feature extraction within the Efficient Channel Attention (ECA) mechanism, forming an M-ECA structure to capture channel features at different levels; a pixel attention mechanism was also introduced to filter out irrelevant or noisy information in the images. Following this, a parameter-free attention module (SimAM) was introduced after deep convolution in the ConvNeXt Block to reweight the input features. SANet, which combines M-ECA and pixel attention mechanisms, was employed at the end of the module to further enhance the model’s dynamic extraction capability of channel and pixel features. Considering the model’s generalization capability, transfer learning was utilized to migrate the pretrained weights of ConvNeXt on the ImageNet dataset to the SA-ConvNeXt model. During training, the Focal Loss function and the Adam optimizer were used to address sample imbalance and reduce gradient fluctuations, thereby enhancing training stability. Finally, the Grad-CAM++ technique was used to generate heatmaps of classification predictions, facilitating the visualization of effective features and deepening the understanding of the model’s focus areas. Comparative experiments were conducted on the Oxford Flowers102 flower image dataset. Compared to existing flower image classification technologies, SA-ConvNeXt performed excellently, achieving a high accuracy of 96.7% and a recall rate of 98.2%, with improvements of 4.0% and 3.7%, respectively, compared to the original ConvNeXt. The results demonstrate that SA-ConvNeXt can effectively capture more accurate key features of flower images, providing an effective technical means for flower recognition and classification.

Robust Commonsense Reasoning Against Noisy Labels Using Adaptive Correction.

Commonsense reasoning based on knowledge graphs (KGs) is a challenging task that requires predicting complex questions over the described textual contexts and relevant knowledge about the world. However, current methods typically assume clean training scenarios with accurately labeled samples, which are often unrealistic. The training set can include mislabeled samples, and the robustness to label noises is essential for commonsense reasoning methods to be practical, but this problem remains largely unexplored. This work focuses on commonsense reasoning with mislabeled training samples and makes several technical contributions: 1) we first construct diverse augmentations from knowledge and model, and offer a simple yet effective multiple-choice alignment method to divide the training samples into clean, semi-clean, and unclean parts; 2) we design adaptive label correction methods for the semi-clean and unclean samples to exploit the supervised potential of noisy information; and 3) finally, we extensively test these methods on noisy versions of commonsense reasoning benchmarks (CommonsenseQA and OpenbookQA). Experimental results show that the proposed method can significantly enhance robustness and improve overall performance. Furthermore, the proposed method is generally applicable to multiple existing commonsense reasoning frameworks to boost their robustness. The code is available at https://github.com/xdxuyang/CR_Noisy_Labels.

Enhancing cross-evidence reasoning graph for document-level relation extraction.

The objective of document-level relation extraction (RE) is to identify the semantic connections that exist between named entities present within a document. However, most entities are distributed among different sentences, there is a need for inter-entity relation prediction across sentences. Existing research has focused on framing sentences throughout documents to predict relationships between entities. However, not all sentences play a substantial role in relation extraction, which inevitably introduces noisy information. Based on this phenomenon, we believe that we can extract evidence sentences in advance and use these evidence sentences to construct graphs to mine semantic information between entities. Thus, we present a document-level RE model that leverages an Enhancing Cross-evidence Reasoning Graph (ECRG) for improved performance. Specifically, we design an evidence extraction rule based on center-sentence to pre-extract higher-quality evidence. Then, this evidence is constructed into evidence graphs to mine the connections between mentions within the same evidence. In addition, we construct entity-level graphs by aggregating mentions from the same entities within the evidence graphs, aiming to capture distant interactions between entities. Experiments result on both DocRED and RE-DocRED datasets demonstrate that our model improves entity RE performance compared to existing work.