With the rapid development of vocational education, there is an increasing demand for intelligent systems that can provide real-time feedback and personalized guidance in practical skills training. Traditional training methods relying on repetitive practice and instructor supervision are inefficient and lack precision. To address these challenges, this paper proposes DualStreamNet (DSNet), an innovative action recognition method designed to enhance the quality of practical teaching in higher vocational education. DSNet integrates spatial and temporal convolutions within a unified framework, enabling simultaneous modeling of skeleton data’s spatial topology and temporal dynamics. Additionally, an uncertainty estimation module is introduced to evaluate prediction credibility, improving the model’s robustness in handling ambiguous actions. Experimental results demonstrate that DSNet achieves superior accuracy and reliability compared to existing methods, providing a solid foundation for developing intelligent feedback systems in vocational education.
PambudiNAHarjantoB. Vocational education in Indonesia: history, development, opportunities, and challenges. Child Youth Serv Rev2020; 115: 105092.
2.
TilakJB. Vocational education and training in Asia. In: International Handbook of Educational Research in the Asia-Pacific Region: Part One. Springer, 2003, pp. 673–686.
3.
KongYFuY. Human action recognition and prediction: a survey. Int J Comput Vis2022; 130: 1366–1401.
4.
WangMXingJMeiJ, et al.Actionclip: adapting language-image pretrained models for video action recognition. IEEE Transact Neural Networks Learn Syst2023.
5.
MazziaVAngaranoSSalvettiF, et al.Action transformer: a self-attention model for short-time pose-based human action recognition. Pattern Recogn2022; 124: 108487.
6.
CongSZhouY. A review of convolutional neural network architectures and their optimizations. Artif Intell Rev2023; 56: 1905–1969.
7.
ParkSSeokC. GalaxyWater-CNN: prediction of water positions on the protein structure by a 3D-convolutional neural network. J Chem Inf Model2022; 62: 3157–3168.
8.
ShahAAMalikHAMMuhammadA, et al.Deep learning ensemble 2D CNN approach towards the detection of lung cancer. Sci Rep2023; 13: 2987.
9.
LockeSBashallAAl-AdelyS, et al.Natural language processing in medicine: a review. Trends Anaesth Crit Care2021; 38: 4–9.
10.
de la ParteMSEMartínez-OrtegaJ-FCastillejoP, et al.Spatio-temporal semantic data management systems for IoT in agiculture 5.0: challenges and future directions. Internet of Things2024; 25: 101030.
11.
FengLZhaoYZhaoW, et al.A comparative review of graph convolutional networks for human skeleton-based action recognition. Artif Intell Rev2022; 55: 1–31.
12.
DurstewitzDKoppeGThurmMI. Reconstructing computational system dynamics from neural data with recurrent neural networks. Nat Rev Neurosci2023; 24: 693–710.
13.
FengZ-kZhangJ-sNiuW-j. A state-of-the-art review of long short-term memory models with applications in hydrology and water resources. Appl Soft Comput2024; 167: 112352.
14.
BhattiUATangHWuG, et al.Deep learning with graph convolutional networks: an overview and latest applications in computational intelligence. Int J Intell Syst2023; 2023: 8342104.
15.
YuanHLiG. A survey of traffic prediction: from spatio-temporal data to intelligent transportation. Data Sci Eng2021; 6: 63–85.
16.
ChenYDingFZhaiL. Multi-scale temporal features extraction based graph convolutional network with attention for multivariate time series prediction. Expert Syst Appl2022; 200: 117011.
17.
DingYZhangZZhaoX, et al.Multi-feature fusion: graph neural network and CNN combining for hyperspectral image classification. Neurocomputing2022; 501: 246–257.
18.
LiuMSunXWangQ, et al.Short-term load forecasting using EMD with feature selection and TCN-based deep learning model. Energies2022; 15: 7170.
19.
TuF-FLiuD-JYanZ-W, et al.STFT-TCAN: a TCN-attention based multivariate time series anomaly detection architecture with time-frequency analysis for cyber-industrial systems. Comput Secur2024; 144: 103961.
20.
HanXZhuXPedryczW, et al.A design of fuzzy rule-based classifier optimized through softmax function and information entropy. Appl Soft Comput2024; 156: 111498.
21.
VonderHLElviraTOchoaO. An analysis of explainability methods for convolutional neural networks. Eng Appl Artif Intell2023; 117: 105606.
22.
MusallamYKAlFassamNIMuhammadG, et al.Electroencephalography-based motor imagery classification using temporal convolutional network fusion. Biomed Signal Process Control2021; 69: 102826.
23.
LodiseTPDrusanoG. Vancomycin area under the curve–guided dosing and monitoring for adult and pediatric patients with suspected or documented serious methicillin-resistant Staphylococcus aureus infections: putting the safety of our patients first. Clin Infect Dis2021; 72: 1497–1501.
24.
KhalifaMAlbadawyM. AI in diagnostic imaging: revolutionising accuracy and efficiency. Computer Methods and Programs in Biomedicine Update2024; 5: 100146.
25.
TakahashiKYamamotoKKuchibaA, et al.Confidence interval for micro-averaged F 1 and macro-averaged F 1 scores. Appl Intell2022; 52: 4961–4972.
26.
LuoRAnandAZulkernineF, et al.YOLO-I3D: optimizing inflated 3D models for real-time human activity recognition. J Imaging2024; 10: 269.
27.
RyuSHongSLeeS. Making TSM better: preserving foundational philosophy for efficient action recognition. ICT Express2024; 10: 570–575.
28.
CuiDXinCWuL, et al.ConvTransformer Attention Network for temporal action detection. Knowl Base Syst2024; 300: 112264.
29.
LiHFanLGaoY, et al.Coupling video vision transformer (ViVit) into land change simulation: a comparison with three-dimensional convolutional neural network (3DCNN). Spatial Sci2024; 69: 873–895.
30.
YangJZhangNZhangJ, et al.Knockdown of MicroRNA160a/b by STTM leads to root architecture changes via auxin signaling in Solanum tuberosum. Plant Physiol Biochem2021; 166: 939–949.
