Gesture recognition in dynamic images is challenging in computer vision, automation and medical field. Hand gesture tracking and recognition between both human and computer must have symmetry in real world. With advances in sensor technology, numerous researchers have recently proposed RGB gesture recognition techniques. In our research paper, we introduce a reliable hand gesture tracking and recognition model that is accurate despite any complex environment, it can track and recognise RGB dynamic gestures. Firstly, videos are converted into frames. After images light intensity adjustment and noise removal, images are passed through CNN for hand gesture extraction. Then from the extracted hand, features are extracted from full hand. Neural gas and locomotion thermal mapping are extracted to make the feature vector. The feature vector are then passed through the fuzzy optimiser to reduce the uncertainties and the fuzziness. The optimised features are then passed to the classifier Deep Belief Network (DBW) for the classification of the gestures. Egogesture and Jester datasets are used for the validation of proposed systems. The experimental results over Egogesture and Jester datasets demonstrate overall accuracies of 90.73% and 89.33% respectively. The experiments proves our system readability and suitability of our proposed model with the other state of the arts model.

Autonomous vehicle detection and tracking are crucial for intelligent transportation management and control systems. Although many techniques are used to develop smart traffic systems, this article discusses vehicle detection and tracking using pixel-labeling and real-time tracking. We propose a novel smart traffic control system that segments the image using an Extreme Gradient Boost (XGBoost) classifier to extract the foreground objects. The proposed model is divided into the following steps: 1) at first, all the images are preprocessed to remove noise; 2) pixel-labeling is performed by using the XGBoost classifier to separate the background from the foreground; 3) all the pixels classified as a vehicle was extracted and converted into a binary image, then blob extraction technique is used to localize each vehicle; 4) to verify the detected vehicles Intersection over Union (IoU) score using the ground truth is calculated; 5) all verified vehicles were subjected to Visual Geometry Group (VGG) feature extraction and based on which a unique identifier was assigned to each of them to enable multi-object tracking across the image frames; 6) vehicles are counted and categorized into stationary and moving cars by detecting motion in each of them using Farneback optical flow algorithm and 6) finally, the Simple Online and Real-time Tracker (SORT) is used for tracking. The proposed model outperforms existing state-of-the-art traffic monitoring techniques in terms of precision, achieving 0.86 for detection and 0.92 for tracking with the Karlsruher Institut for Technology Aerial Image Sequences (KIT-AIS) dataset, 0.83 for detection, and 0.87 for tracking with the Vision Meets Drone Single Object-Tracking (VisDrone) dataset. The proposed system can be used for several purposes, such as vehicle identification in traffic, traffic density detection at intersections, traffic flow conditions on the road, and providing a pedestrian way.

The Soyang Dam, the largest multipurpose dam in Korea, faces water resource management challenges due to global warming. Global warming increases the duration and frequency of days with high temperatures and extreme precipitation events. Therefore, it is crucial to accurately predict the inflow rate for water resource management because it helps plan for flood, drought, and power generation in the Seoul metropolitan area. However, the lack of hydrological data for the Soyang River Dam causes a physical-based model to predict the inflow rate inaccurately. This study uses nearly 15 years of meteorological, dam, and weather warning data to overcome the lack of hydrological data and predict the inflow rate over two days. In addition, a sequence-to-sequence (Seq2Seq) mechanism combined with a bidirectional long short-term memory (LSTM) is developed to predict the inflow rate. The proposed model exhibits state-of-the-art prediction accuracy with root mean square error (RMSE) of 44.17 m³/s and 58.59 m³/s, mean absolute error (MAE) of 14.94 m³/s and 17.11 m³/s, and Nash–Sutcliffe efficiency (NSE) of 0.96 and 0.94, for forecasting first and second day, respectively.

Human-human interaction recognition is crucial in computer vision fields like surveillance, human-computer interaction, and social robotics. It enhances systems’ ability to interpret and respond to human behavior precisely. This research focuses on recognizing human interaction behaviors using a static image, which is challenging due to the complexity of diverse actions. The overall purpose of this study is to develop a robust and accurate system for human interaction recognition. This research presents a novel image-based human interaction recognition method using a Hidden Markov Model (HMM). The technique employs hue, saturation, and intensity (HSI) color transformation to enhance colors in video frames, making them more vibrant and visually appealing, especially in low-contrast or washed-out scenes. Gaussian filters reduce noise and smooth imperfections followed by silhouette extraction using a statistical method. Feature extraction uses the features from Accelerated Segment Test (FAST), Oriented FAST, and Rotated BRIEF (ORB) techniques. The application of Quadratic Discriminant Analysis (QDA) for feature fusion and discrimination enables high-dimensional data to be effectively analyzed, thus further enhancing the classification process. It ensures that the final features loaded into the HMM classifier accurately represent the relevant human activities. The impressive accuracy rates of 93% and 94.6% achieved in the BIT-Interaction and UT-Interaction datasets respectively, highlight the success and reliability of the proposed technique. The proposed approach addresses challenges in various domains by focusing on frame improvement, silhouette and feature extraction, feature fusion, and HMM classification. This enhances data quality, accuracy, adaptability, reliability, and reduction of errors.

Most recommenders generate recommendations for a user by computing the preference score of items, sorting the items according to the score, and filtering top- $K$ -items of high scores. Since sorting is not differentiable and is difficult to optimize with gradient descent, it is nontrivial to incorporate it in recommendation model training despite its relevance to top- $K$ recommendations. As a result, inconsistency occurs between existing learning objectives and ranking metrics of recommenders. In this work, we present the Differentiable Ranking Metric (DRM) that mitigates the inconsistency between model training and generating top- $K$ recommendations, aiming at improving recommendation performance by employing the differentiable relaxation of ranking metrics via joint learning. Using experiments with several real-world datasets, we demonstrate that the joint learning of the DRM objective and existing factor based recommenders significantly improves the quality of recommendations.

The adoption of recommender systems in business-to-business (B2B) can make the management of companies more efficient. Although the importance of recommendation is increasing with the expansion of B2B e-commerce, not enough studies on B2B recommendations have been conducted. Due to several differences between B2B and business-to-consumer (B2C), the B2B recommender system should be defined differently. This paper presents a new perspective on the explainable B2B recommender system using the knowledge graph attention network for recommendation (KGAT). Unlike traditional recommendation systems that suggest products to consumers, this study focuses on recommending potential buyers to sellers. Additionally, the utilization of the KGAT attention mechanisms enables the provision of explanations for each company’s recommendations. The Korea Electronic Taxation System Association provides the Market Transaction Dataset in South Korea, and this research shows how the dataset is utilized in the knowledge graph (KG). The main tasks can be summarized in three points: (i) suggesting the application of an explainable recommender system in B2B for recommending potential customers, (ii) extracting the performance-enhancing features of a knowledge graph, and (iii) enhancing keyword extraction for trading items to improve recommendation performance. We can anticipate providing good insight into the development of the industry via the utilization of the B2B recommendation of potential customer prediction.