Kaoru Hirota’s research while affiliated with Beijing Institute of Technology and other places

A multiscale sequence information fusion (MSSIF) method is presented for dynamic facial expression recognition (DFER) in video sequences. It exploits multiscale information by integrating features from individual frames, subsequences, and entire sequences through a transformer-based architecture. This hierarchical feature fusion process includes deep feature extraction at the frame level to capture intricate visual details, intrasubsequence fusion using self-attention mechanisms for analyzing adjacent frames, and intersubsequence fusion to synthesize long-term emotional dynamics across time scales. The efficacy of MSSIF is demonstrated through extensive evaluation on three video datasets: eNTERFACE’05, BAUM-1s, and AFEW, where it achieves overall recognition accuracies of 60.1%, 60.7%, and 58.8%, respectively. These results substantiate MSSIF’s superior performance in accurately recognizing facial expressions by managing short and long-term dependencies within video sequences, making it a potent tool for real-world applications requiring nuanced dynamic facial expression detection.

Accurate determination of volume percentages in three-phase fluids is paramount for the success of various industrial processes, ranging from oil and gas production to chemical engineering. This study presents a comprehensive approach to this challenge by leveraging advanced signal processing techniques and machine learning paradigms. Our methodology integrates the time, frequency, and wavelet transform features extracted from X-ray-based measurement systems whose structure consists of an X-ray tube source, two sodium iodide detectors, and a test pipe, all of which were simulated using the Monte Carlo N Particle code. The amalgamation of these features provides a rich representation of the fluid composition that captures both temporal and spectral characteristics. To enhance the discriminative power of the features, we employ a simulated annealing algorithm to strategically reduce their dimensionality and select pertinent features. The simulated annealing unit systematically evaluates the contribution of each feature to predictive accuracy. Further, through iterative elimination and re-evaluation, the algorithm refines the feature set, retaining only those with the highest relevance to the three-phase fluid composition. This feature selection process optimises the performance of subsequent machine learning models, streamlining the input space for enhanced interpretability and efficiency. Finally, to determine the volume percentages, we employ a support vector regression (SVR) neural network, which is trained on a refined dataset with capability to handle complex relationships and high-dimensional data. The proposed approach demonstrates superior accuracy in determining volume percentages of three-phase fluids compared to traditional methods, thereby making it an effective and integrated technique to analyse fluid composition in a variety of industrial settings and applications.

To strengthen the population diversity and search capability of equilibrium optimizer (EO), a dynamic multi-population mutation architecture-based equilibrium optimizer (DMMAEO) is proposed. Firstly, a dynamic multi-population guidance mechanism is constructed to enhance population diversity. Secondly, a dynamic Gaussian mutation-based sub-population concentration updating mechanism is introduced to strengthen exploitation ability. Finally, a dynamic Cauchy mutation-based sub-population equilibrium candidate generation mechanism is integrated to boost exploration ability. The optimization ability of DMMAEO is assessed through a comparison with several recent promising algorithms on 58 test functions (including 29 representative test functions and 29 CEC2017 test functions). The comparison results reveal that the DMMAEO has superiority in the performance assessment of seeking global optimum over other compared algorithms. The DMMAEO is further employed in addressing six engineering design problems and a UGV multi-target path planning problem. The results show the practicality of DMMAEO in addressing engineering application tasks. The aforementioned numerical optimization and engineering application experimental results show that the three enhancement mechanisms of DMMAEO improve the optimization ability of the canonical EO, and the DMMAEO has competitiveness in tackling various kinds of complex numerical optimization and engineering application problems.