Intelligent Data Analysis: An International Journal

Capturing images through semi-reflective surfaces, such as glass windows and transparent enclosures, often leads to a reduction in visual quality and can adversely affect the performance of computer vision algorithms. As a result, image reflection removal has garnered significant attention among computer vision researchers. With the growing application of deep learning methods in various computer vision tasks, such as super-resolution, inpainting, and denoising, convolutional neural networks (CNNs) have become an increasingly popular choice for image reflection removal. The purpose of this paper is to provide a comprehensive review of learning-based algorithms designed for image reflection removal. Firstly, we provide an overview of the key terminology and essential background concepts in this field. Next, we examine various datasets and data synthesis methods to assist researchers in selecting the most suitable options for their specific needs and targets. We then review existing methods with qualitative and quantitative results, highlighting their contributions and significance in this field. Finally, some considerations about challenges and future scope in image reflection removal techniques are discussed.

The quest for transparency in black-box models has gained significant momentum in recent years. In particular, discovering the underlying machine learning technique type (or model family) from the performance of a black-box model is a real important problem both for better understanding its behaviour and for developing strategies to attack it by exploiting the weaknesses intrinsic to the learning technique. In this paper, we tackle the challenging task of identifying which kind of machine learning model is behind the predictions when we interact with a black-box model. Our innovative method involves systematically querying a black-box model (oracle) to label an artificially generated dataset, which is then used to train different surrogate models using machine learning techniques from different families (each one trying to partially approximate the oracle’s behaviour). We present two approaches based on similarity measures, one selecting the most similar family and the other using a conveniently constructed meta-model. In both cases, we use both crisp and soft classifiers and their corresponding similarity metrics. By experimentally comparing all these methods, we gain valuable insights into the explanatory and predictive capabilities of our model family concept. This provides a deeper understanding of the black-box models and increases their transparency and interpretability, paving the way for more effective decision making.

Real-world machine learning applications seldom provide perfect labeled data, posing a challenge in developing models robust to noisy labels. Recent methods prioritize noise filtering based on the discrepancies between model predictions and the provided noisy labels, assuming samples with minimal classification losses to be clean. In this work, we capitalize on the consistency between the learned model and the complete noisy dataset, employing the data’s rich representational and topological information. We introduce LaplaceConfidence, a method that to obtain label confidence (i.e., clean probabilities) utilizing the Laplacian energy. Specifically, it first constructs graphs based on the feature representations of all noisy samples and minimizes the Laplacian energy to produce a low-energy graph. Clean labels should fit well into the low-energy graph while noisy ones should not, allowing our method to determine data’s clean probabilities. Furthermore, LaplaceConfidence is embedded into a holistic method for robust training, where co-training technique generates unbiased label confidence and label refurbishment technique better utilizes it. We also explore the dimensionality reduction technique to accommodate our method on large-scale noisy datasets. Our experiments demonstrate that LaplaceConfidence outperforms state-of-the-art methods on benchmark datasets under both synthetic and real-world noise. Code available at https://github.com/chenmc1996/LaplaceConfidence.

Traffic flow prediction can improve transportation efficiency, which is an important part of intelligent transportation systems. In recent years, the prediction method based on graph convolutional recurrent neural network has been widely used in traffic flow prediction. However, in real application scenarios, the spatial dependence of graph signals will change with time, and the filter using a fixed graph displacement operator cannot accurately predict traffic flow at the current moment. To improve the accuracy of traffic flow prediction, a two-layer graph convolutional recurrent neural network based on the dynamic graph displacement operator is proposed. The framework of our proposal is to use the first layer of static graph convolutional recurrent neural network to generate the sequence wave vector of the graph displacement operator. The sequence wave vector is passed through the deconvolutional neural network to obtain the sequence dynamic graph displacement operator, and then the second layer dynamic graph convolutional recurrent neural network is used to predict the traffic flow at the next moment. The model is evaluated on the METR-LA and PEMS-BAY datasets. Experimental results demonstrate that our model significantly outperforms other baseline models.

In the process of product ranking considering online reviews, they often are based on initial reviews and do not consider additional consumer reviews, but additional review information can sometimes directly affect consumers’ final decisions. To fully characterize the rich emotional preferences of consumers embedded in two-stage online customer reviews information, considering consumers’ individual preferences and product objective evaluation information, we construct a combination weighting method to calculate comprehensive weights of product attributes, and then exploit the sentiment analysis technique, interval-valued probabilistic linguistic term set (IVPLTS) and preference ranking organization method for enrichment evaluations (PROMETHEE) to establish a products ranking method based on compound reviews, and then we use it to identify the sentiment orientation of reviews and the results. Finally, a real-life case illustrates a real-world application of the proposed method.

While time series data are prevalent across diverse sectors, data labeling process still remains resource-intensive. This results in a scarcity of labeled data for deep learning, emphasizing the importance of semi-supervised learning techniques. Applying semi-supervised learning to time series data presents unique challenges due to its inherent temporal complexities. Efficient contrastive learning for time series requires specialized methods, particularly in the development of tailored data augmentation techniques. In this paper, we propose a single-step, semi-supervised contrastive learning framework named nearest neighbor contrastive learning for time series (NNCLR-TS). Specifically, the proposed framework incorporates a support set to store representations including their label information, enabling a pseudo-labeling of the unlabeled data based on nearby samples in the latent space. Moreover, our framework presents a novel data augmentation method, which selectively augments only the trend component of the data, effectively preserving their inherent periodic properties and facilitating effective training. For training, we introduce a novel contrastive loss that utilizes the nearest neighbors of augmented data for positive and negative representations. By employing our framework, we unlock the ability to attain high-quality embeddings and achieve remarkable performance in downstream classification tasks, tailored explicitly for time series. Experimental results demonstrate that our method outperforms the state-of-the-art approaches across various benchmarks, validating the effectiveness of our proposed method.

In order to alleviate urban congestion, improve vehicle mobility, and improve logistics delivery efficiency, this paper establishes a practical multi-objective and multi constraint logistics delivery mathematical model based on graphs, and proposes a solution algorithm framework that combines decomposition strategy and deep reinforcement learning (DRL). Firstly, taking into account the actual multiple constraints such as customer distribution, vehicle load constraints, and time windows in urban logistics distribution regions, a multi constraint and multi-objective urban logistics distribution mathematical model was established with the goal of minimizing the total length, cost, and maximum makespan of urban logistics distribution paths. Secondly, based on the decomposition strategy, a DRL framework for optimizing urban logistics delivery paths based on Graph Capsule Network (G-Caps Net) was designed. This framework takes the node information of VRP as input in the form of a 2D graph, modifies the graph attention capsule network by considering multi-layer features, edge information, and residual connections between layers in the graph structure, and replaces probability calculation with the module length of the capsule vector as output. Then, the baseline REINFORCE algorithm with rollout is used for network training, and a 2-opt local search strategy and sampling search strategy are used to improve the quality of the solution. Finally, the performance of the proposed method was evaluated on standard examples of problems of different scales. The experimental results showed that the constructed model and solution framework can improve logistics delivery efficiency. This method achieved the best comprehensive performance, surpassing the most advanced distress methods, and has great potential in practical engineering.

Air pollution is an alarming problem in many cities and countries around the globe. The ability to forecast air pollutant levels plays a crucial role in implementing necessary prevention measures to curb its effects in advance. There are many statistical, machine learning, and deep learning models available to predict air pollutant values, but only a limited number of models take into account the spatio-temporal factors that influence pollution. In this study a novel Deep Learning model that is augmented with Spatio-Temporal Co-Occurrence Patterns (STEEP) is proposed. The deep learning model uses the Closed Spatio-Temporal Co-Occurrence Pattern mining (C-STCOP) algorithm to extract non-redundant/closed patterns and the Diffusion Convolution Recurrent Neural Network (DCRNN) for time series prediction. By constructing a graph based on the co-occurrence patterns obtained from C-STCOP, the proposed model effectively addresses the spatio-temporal association among monitoring stations. Furthermore, the sequence-to-sequence encoder-decoder architecture captures the temporal dependencies within the time series data. The STEEP model is evaluated using the Delhi air pollutants dataset and shows an average improvement of 8%–13% in RMSE, MAE and MAPE metric compared to the baseline models.

The Usage-Based Insurance paradigm, which is receiving a lot of attention in recent years, envisages computing the car policy premium based on accident risk probability, evaluated observing the past driving history and habits. However, Usage-Based Insurance strategies are usually based on simple empirical decision rules built on travelled distance. The development of intelligent systems for smart risk prediction using the stored overall driving behaviour, without the need of other insurance or socio-demographic information, is still an open challenge. This work aims at exploring a comprehensive machine learning-based approach solely based on driving-related data of private vehicles. The anonymized dataset employed in this study is provided by the telematics company UnipolTech, and contains space/time densely measured data related to trips of almost 100000 vehicles uniformly spread on the Italian territory, recorded every 2 km by on-board telematics fix devices (black boxes), from February 2018 to February 2020. An innovative feature engineering process is proposed, with the aim of uncovering novel informative quantities able to disclose complex aspects of driving behaviour. Recent and powerful learning techniques are explored to develop advanced predictive models, able to provide a reliable accident probability for each vehicle, automatically managing the critical imbalance intrinsically peculiar this kind of datasets.

The emergence of the novel COVID-19 virus has had a profound impact on global healthcare systems and economies, underscoring the imperative need for the development of precise and expeditious diagnostic tools. Machine learning techniques have emerged as a promising avenue for augmenting the capabilities of medical professionals in disease diagnosis and classification. In this research, the EFS-XGBoost classifier model, a robust approach for the classification of patients afflicted with COVID-19 is proposed. The key innovation in the proposed model lies in the Ensemble-based Feature Selection (EFS) strategy, which enables the judicious selection of relevant features from the expansive COVID-19 dataset. Subsequently, the power of the eXtreme Gradient Boosting (XGBoost) classifier to make precise distinctions among COVID-19-infected patients is harnessed.The EFS methodology amalgamates five distinctive feature selection techniques, encompassing correlation-based, chi-squared, information gain, symmetric uncertainty-based, and gain ratio approaches. To evaluate the effectiveness of the model, comprehensive experiments were conducted using a COVID-19 dataset procured from Kaggle, and the implementation was executed using Python programming. The performance of the proposed EFS-XGBoost model was gauged by employing well-established metrics that measure classification accuracy, including accuracy, precision, recall, and the F1-Score. Furthermore, an in-depth comparative analysis was conducted by considering the performance of the XGBoost classifier under various scenarios: employing all features within the dataset without any feature selection technique, and utilizing each feature selection technique in isolation. The meticulous evaluation reveals that the proposed EFS-XGBoost model excels in performance, achieving an astounding accuracy rate of 99.8%, surpassing the efficacy of other prevailing feature selection techniques. This research not only advances the field of COVID-19 patient classification but also underscores the potency of ensemble-based feature selection in conjunction with the XGBoost classifier as a formidable tool in the realm of medical diagnosis and classification.

At the core of Deep Learning-based Deformable Medical Image Registration (DMIR) lies a strong foundation. Essentially, this network compares features in two images to identify their mutual correspondence, which is necessary for precise image registration. In this paper, we use three novel techniques to increase the registration process and enhance the alignment accuracy between medical images. First, we propose cross attention over multi-layers of pairs of images, allowing us to take out the correspondences between them at different levels and improve registration accuracy. Second, we introduce a skip connection with residual blocks between the encoder and decoder, helping information flow and enhancing overall performance. Third, we propose the utilization of cascade attention with residual block skip connections, which enhances information flow and empowers feature representation. Experimental results on the OASIS data set and the LPBA40 data set show the effectiveness and superiority of our proposed mechanism. These novelties contribute to the enhancement of 3D DMIR-based on unsupervised learning with potential implications in clinical practice and research.

Content-Based Image Retrieval (CBIR) uses complicated algorithms to analyze visual attributes and retrieve relevant photos from large databases. CBIR is essential to a privacy-preserving feature extraction and protection method for outsourced picture representation. SecureImageSec combines essential methods with the system’s key entities to ensure secure, private and protected image feature processing during outsourcing. For a system to be implemented effectively, these techniques must be seamlessly integrated across critical entities, such as the client, the cloud server that is being outsourced, the component that protects secure features, the component that maintains privacy in communication, access control, and authorization, and the integration and system evaluation. The client entity initiates outsourcing using advanced encryption techniques to protect privacy. SecureImageSec protects outsourced data by using cutting-edge technologies like Fully Homomorphic Encryption (FHE) and Secure Multi-Party Computation (SMPC). Cloud servers hold secure feature protection entities and protect outsourced features’ privacy and security. SecureImageSec uses AES and FPE to protect data format. SecureImageSec’s cloud-outsourced privacy-preserving communication uses SSL/TLS and QKD to protect data transmission. Attribute-Based Encryption (ABE) and Functional Encryption (FE) in SecureImageSec limit access to outsourced features based on user attributes and allow fine-grained access control over decrypted data. SecureImageSec’s Information Leakage Rate (ILR) of 0.02 for a 1000-feature dataset shows its efficacy. SecureImageSec also achieves 4.5 bits of entropy, ensuring the encrypted feature set’s muscular cryptographic strength and randomness. Finally, SecureImageSec provides secure and private feature extraction and protection, including CBIR capabilities, for picture representation outsourcing.

Video grounding intends to perform temporal localization in multimedia information retrieval. The temporal bounds of the target video span are determined for the given input query. A novel interactive multi-head self-attention (IMSA) transformer is proposed to localize an unseen moment in the untrimmed video for the given image. A new semantic-trained self-supervised approach is considered in this paper to perform cross-domain learning to match the image query – video segment. It normalizes the convolution function enabling efficient correlation and collecting of semantically related video segments across time based on the image query. A double hostile Contrastive learning with Gaussian distribution parameters method is advanced to learn the representations of video. The proposed approach performs dynamically on various video components to achieve exact semantic synchronization and localization among queries and video. In the proposed approach, the IMSA model localizes frames greatly compared to other approaches. Experiments on benchmark datasets show that the proposed model can significantly increase temporal grounding accuracy. The moment occurrence is identified in the video with a start and end boundary ascertains an average recall of 86.45% and a mAP of 59.3%.

With the rapid development of unmanned aerial vehicle (UAV) technology and computer vision, real-time object detection in UAV aerial images has become a current research hotspot. However, the detection tasks in UAV aerial images face challenges such as disparate object scales, numerous small objects, and mutual occlusion. To address these issues, this paper proposes the ASM-YOLO model, which enhances the original model by replacing the Neck part of YOLOv8 with an efficient bidirectional cross-scale connections and adaptive feature fusion (ABiFPN) . Additionally, a Structural Feature Enhancement Module (SFE) is introduced to inject features extracted by the backbone network into the Neck part, enhancing inter-network information exchange. Furthermore, the MPDIoU bounding box loss function is employed to replace the original CIoU bounding box loss function. A series of experiments was conducted on the VisDrone-DET dataset, and comparisons were made with the baseline network YOLOv8s. The experimental results demonstrate that the proposed model in this study achieved reductions of 26.1% and 24.7% in terms of parameter count and model size, respectively. Additionally, during testing on the evaluation set, the proposed model exhibited improvements of 7.4% and 4.6% in the AP50 and mAP metrics, respectively, compared to the YOLOv8s baseline model, thereby validating the practicality and effectiveness of the proposed model. Subsequently, the generalizability of the algorithm was validated on the DOTA and DIOR datasets, which share similarities with aerial images captured by drones. The experimental results indicate significant enhancements on both datasets.