Hybrid Human Activity Recognition: Integrating Traditional Feature Engineering with Deep Learning Approach

Human Activity Recognition (HAR) is a vital research area with applications in healthcare, security, and intelligent environments. This paper presents a hybrid framework that combines traditional feature engineering with deep learning to enhance HAR performance. It leverages the Histogram of Oriented Gradients (HoG) for spatial feature extraction and Support Vector Machines (SVM) for structured classification. Additionally, Vision Transformers (ViT) and ResNet architectures are integrated to improve accuracy: ViT captures global dependencies through attention mechanisms, while ResNet enhances deep feature learning through skip connections. Experimental results demonstrate that this approach balances computational efficiency, interpretability, and high accuracy on large datasets.

Human Activity Recognition (HAR)

1. Khan, Irfanullah, Antonio Guerrieri, Edoardo Serra, and Giandomenico Spezzano. ”A hybrid deep learning model for UWB radar-based human activity recognition.” Internet of Things 29 (2025): 101458. [Google Scholar] [Crossref]

2. Yadav, Santosh Kumar, Kamlesh Tiwari, Hari Mohan Pandey, and Shaik Ali Akbar. A review of multimodal human activity recognition with special emphasis on classification, applications, challenges and future directions. Knowledge- Based Systems 223 (2021): 106970. [Google Scholar] [Crossref]

3. Nawal, Yala, Mourad Oussalah, Belkacem Fer- gani, and Anthony Fleury. New incremental SVM algorithms for human activity recognition in smart homes. Journal of Ambient Intelli- gence and Humanized Computing 14, no. 10 (2023): 13433-13450. [Google Scholar] [Crossref]

4. Athavale, Vijay Anant, Suresh Chand Gupta, Deepak Kumar, and Savita Savita. Human action recognition using CNN-SVM model. Advances in Science and Technology 105 (2021): 282-290. [Google Scholar] [Crossref]

5. Patel, Chirag I., Dileep Labana, Sharnil Pandya, Kirit Modi, Hemant Ghayvat, and Muhammad Awais. Histogram of oriented gradient-based fusion of features for human action recognition in action video sequences. Sensors 20, no. 24 (2020): 7299. [Google Scholar] [Crossref]

6. Maheswari, B. Uma, R. Sonia, M. P. Rajaku- mar, and J. Ramya. Novel machine learning for human actions classification using histogram of oriented gradients and sparse representation. Information Technology and Control 50, no. 4 (2021): 686-705. [Google Scholar] [Crossref]

7. Khan, Zanobya N., and Jamil Ahmad. Atten- tion induced multi-head convolutional neu- ral network for human activity recognition. Applied soft computing 110 (2021): 107671. [Google Scholar] [Crossref]

8. Khan, Saad Irfan, Hussain Dawood, M. A. Khan, Ghassan F. Issa, Amir Hussain, Mrim M. Alnfiai, and Khan Muhammad Adnan. ”Transition-aware human activity recognition using an ensemble deep learning framework.” Computers in Human Behavior 162 (2025):108435. [Google Scholar] [Crossref]

9. Wensel, James, Hayat Ullah, and Arslan Munir. Vit-ret: Vision and recurrent transformer neu- ral networks for human activity recognition in videos. IEEE Access 11 (2023): 72227-72249. [Google Scholar] [Crossref]

10. Qu, Lele, Xiayang Li, Tianhong Yang, and Shuang Wang. ”Radar-Based Continuous Human Activity Recognition Using Multido main Fusion Vision Transformer.” IEEE Sen sors Journal (2025). [Google Scholar] [Crossref]

11. Mekruksavanich, Sakorn, and Anuchit Jitpat tanakul. Lstm networks using smartphone data for sensor-based human activity recognition in smart homes. Sensors 21, no. 5 (2021): 1636. [Google Scholar] [Crossref]

12. Ankita, Shalli Rani, Himanshi Babbar, Sonya Coleman, Aman Singh, and Hani Moaiteq Aljahdali. An efficient and lightweight deep learning model for human activity recognition using smartphones. Sensors 21, no. 11 (2021): 3845. [Google Scholar] [Crossref]

13. Zhang, Shibo, Yaxuan Li, Shen Zhang, Farzad Shahabi, Stephen Xia, Yu Deng, and Nabil Alshurafa. Deep learning in human activity recognition with wearable sensors: A review on advances. Sensors 22, no. 4 (2022): 1476. [Google Scholar] [Crossref]

14. Xu, Yang, and Ting Ting Qiu. Human activity recognition and embedded application based on convolutional neural network. Journal of Artifi cial Intelligence and Technology 1, no. 1 (2021): 51-60. [Google Scholar] [Crossref]

15. Khan, Muhammad Attique, Kashif Javed, Sajid Ali Khan, Tanzila Saba, Usman Habib, Junaid Ali Khan, and Aaqif Afzaal Abbasi. Human action recognition using fusion of mul tiview and deep features: an application to video surveillance. Multimedia tools and appli cations 83, no. 5 (2024): 14885-14911. [Google Scholar] [Crossref]

16. Dua, Nidhi, Shiva Nand Singh, and Vijay Bhaskar Semwal. Multi-input CNN-GRU based human activity recognition using wearable sen sors. Computing 103, no. 7 (2021): 1461-1478. [Google Scholar] [Crossref]

17. Muhammad, Khan, Amin Ullah, Ali Shariq Imran, Muhammad Sajjad, Mustafa Servet Kiran, Giovanna Sannino, and Victor Hugo C. de Albuquerque. Human action recogni tion using attention based LSTM network with dilated CNN features. Future Generation Com puter Systems 125 (2021): 820-830.. [Google Scholar] [Crossref]

18. Xiao, Zhiwen, Xin Xu, Huanlai Xing, Fuhong Song, Xinhan Wang, and Bowen Zhao. A federated learning system with enhanced fea ture extraction for human activity recognition. Knowledge-Based Systems 229 (2021): 107338. [Google Scholar] [Crossref]

19. Al-Qaness, Mohammed AA, Abdelghani Dahou, Mohamed Abd Elaziz, and A. M. Helmi. ”Multi-ResAtt: Multilevel residual network with attention for human activity recognition using wearable sensors.” IEEE Transactions on Industrial Informatics 19, no. 1 (2022): 144-152. [Google Scholar] [Crossref]

20. Wang, Jindong, Yiqiang Chen, Shuji Hao, Xiaohui Peng, and Lisha Hu. ”Deep learning for sensor-based activity recognition: A survey.” Pattern recognition letters 119 (2019): 3-11. [Google Scholar] [Crossref]

21. Hussain, Altaf, Tanveer Hussain, Waseem Ullah, and Sung Wook Baik. ”Vision trans former and deep sequence learning for human activity recognition in surveillance videos.” Computational Intelligence and Neuroscience 2022, no. 1 (2022): 3454167. [Google Scholar] [Crossref]

22. Trujillo-Guerrero, Mar´ıa Fernanda, Stadyn Rom´an-Niemes, Milagros Ja´en-Vargas, Alfonso Cadiz, Ricardo Fonseca, and Jos´e Javier Serrano-Olmedo. ”Accuracy comparison of CNN, LSTM, and transformer for activity recognition using IMU and visual markers.” IEEE Access 11 (2023): 106650-106669. [Google Scholar] [Crossref]

23. Ronald, Mutegeki, Alwin Poulose, and Dong Seog Han. ”iSPLInception: an inception ResNet deep learning architecture for human activity recognition.” IEEE Access 9 (2021): 68985-69001. [Google Scholar] [Crossref]

24. Tang, and Aiguo Song. ”Triple cross-domain attention on human activity recognition using wearable sensors.” IEEE Transactions on Emerging Topics in Computational Intelligence 6, no. 5 (2022): 1167-1176. [Google Scholar] [Crossref]

25. Tang, and Jun He. ”Multiscale deep fea ture learning for human activity recognition using wearable sensors.” IEEE Transactions on Industrial Electronics 70, no. 2 (2022): 2106 2116. [Google Scholar] [Crossref]

26. Lin, Xiao, et al. ”EAPT: efficient attention pyramid transformer for image processing.” IEEE Trans on Multimedia 25 (2021): 50-61. [Google Scholar] [Crossref]

27. Yao, Jieteng, et al. ”Scene-aware human pose generation using transformer.” Proceedings of the 31st ACM International Conference on Multimedia. 2023. http://vision.stanford.edu/Datasets/40actions.html [Google Scholar] [Crossref]

• What the Desert Fathers Teach Data Scientists: Ancient Ascetic Principles for Ethical Machine-Learning Practice
• Comparative Analysis of Some Machine Learning Algorithms for the Classification of Ransomware
• Comparative Performance Analysis of Some Priority Queue Variants in Dijkstra’s Algorithm
• Transfer Learning in Detecting E-Assessment Malpractice from a Proctored Video Recordings.
• Dual-Modal Detection of Parkinson’s Disease: A Clinical Framework and Deep Learning Approach Using NeuroParkNet