Skeleton-based human action recognition is an important research area with many practical applications. Most existing methods rely on single representations of skeletal sequences, which cannot totally obtain all the complex features of human movements. This paper presents LFHAR (Latent Features for Human Action Recognition), a new framework that uses multiple spatio-temporal latent representations to improve the extraction of action features. Our method captures how skeletal poses change over time and combines motion information from both individual joints and connected body parts. The proposed approach applies graph-based processing to each skeleton frame in a sequence, then arranges the resulting graph features into spatio-temporal matrices. We tested LFHAR on standard datasets to demonstrate its stability and effectiveness. Our method shows significant improvements, achieving 2.7% higher accuracy on NTU-RGB+D 60 and 2.1% higher accuracy on NTU-RGB+D 120 compared to baseline methods. These results confirm that the LFHAR framework effectively improves skeleton-based action recognition performance.
| Mualliflar | Kabul Khudaybergenov, Avazjon Marakhimov, Zahriddin Muminov |
|---|---|
| Jurnal | Кимёвий технология. Назорат ва бошқарув |
| Nashr sanasi | 2025-11-01 |
| Jild | 2025 |
| Son | 5 |
| Betlar | 88-95 |
| Til | en |
| DOI | 10.59048/2181-1105.1724 |
DOI: 10.59048/2181-1105.1724 · Maqolaning asl sahifasi
This paper examines the stages of multilayer neural network technology for the control and identification of dynamic systems. It is seen many ways to use multilayer neural networks in control systems. Neural controllers…
Intensive development and expansion of the field of application of modern observation systems are largely determined by the improvement of methods and means of interpreting the primary results recorded by the system…
The article analyzes the potential of quantum algorithms for optimizing cognitive decision-making systems in production processes in the energy sector. Compared with traditional algorithms, quantum algorithms allow…
This article discusses the process of monitoring information security incidents and factors affecting the effectiveness of monitoring. A procedural model of the functioning of the information security incident…
This article is devoted to the study of the modeling process of analog-to-digital converters (ADCs) that process signals, one of the main parts of control system elements and devices. As we know, ADCs are an important…
The technological method of storing and processing oilseed raw materials is structurally modelled in the article. Additionally, it is suggested that the generalized structure be presented as a three-part scheme…
The work is devoted to the issue of increasing the accuracy of liquid flow measurement devices, containing a primary measuring transducer for measuring the pressure drop in a narrowed section of a pipeline, temperature…
In this manuscripts, the synthesis of control systems with multilayer neural networks based on the speed gradient methods is given. For adjusting the weight coefficients of the base processor element, the gradient…
The method of controlling the rectification process using fuzzy control based on fuzzy logic is studied. The structural scheme of the rectification process control system based on fuzzy logic is proposed. The analysis…
Presents effective methods and solutions to eliminate errors caused by the “human factor” (fatigue, neglect, etc.) as a result of the creation and improvement of automated systems based on artificial intelligence…