Adaptive Feedback-Driven Segmentation for Continuous Multi-Label Human Activity Recognition
Source abstract
Abstract
Radar-based continuous human activity recognition (HAR) in realistic scenarios faces challenges in segmenting and classifying overlapping or concurrent activities. This paper introduces a feedback-driven adaptive segmentation framework for multi-label classification in continuous HAR, leveraging Bayesian optimization (BO) and reinforcement learning (RL) to dynamically adjust segmentation parameters such as segment length and overlap in the data stream, optimizing them based on performance metrics such as accuracy and F1-score. Using a public dataset of continuous human activities, the method trains ResNet18 models on spectrogram, range-Doppler, and range-time representations from a 20% computational subset. Then, it scales optimized parameters to the full dataset. Comparative analysis against fixed-segmentation baselines was made. The results demonstrate significant improvements in classification performance, confirming the potential of adaptive segmentation techniques in enhancing the accuracy and efficiency of continuous multi-label HAR systems.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Anomaly Detection Techniques and Applications · Human Pose and Action Recognition · Context-Aware Activity Recognition Systems
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Electronic Information Engineering, Beihang University, Beijing 100191, China
Funding
Aeronautical Science Foundation of China
References (32)
Show source-provided references
1. Fioranelli, F., and Le Kernec, J. (November, January 31). Radar sensing for human healthcare: Challenges and results. Proceedings of the 2021 IEEE Sensors, Virtual.10.1109/SENSORS47087.2021.9639702 ↗
2. Passive Radar Sensing for Human Activity Recognition: A Survey10.1109/OJEMB.2024.3420747 ↗
3. Islam, M.M., Nooruddin, S., Karray, F., and Muhammad, G. (2022). Human activity recognition using tools of convolutional neural networks: A state of the art review, data sets, challenges, and future prospects. Comput. Biol. Med., 149.10.1016/j.compbiomed.2022.106060 ↗
4. Human Behavior Analysis Using Radar Data: A Survey10.1109/ACCESS.2024.3474100 ↗
5. Papadopoulos, K., and Jelali, M. (2023). A Comparative Study on Recent Progress of Machine Learning-Based Human Activity Recognition with Radar. Appl. Sci., 13.10.20944/preprints202311.1347.v1 ↗
6. A Survey on Radar-Based Continuous Human Activity Recognition10.1109/JMW.2023.3264494 ↗
7. Radar-Based Human Activity Recognition Using Hybrid Neural Network Model With Multidomain Fusion10.1109/TAES.2021.3068436 ↗
8. Bi-LSTM Network for Multimodal Continuous Human Activity Recognition and Fall Detection10.1109/JSEN.2019.2946095 ↗
9. Ullmann, I., Guendel, R.G., Kruse, N.C., Fioranelli, F., and Yarovoy, A. (November, January 29). Radar-Based Continuous Human Activity Recognition with Multi-Label Classification. Proceedings of the 2023 IEEE Sensors, Vienna, Austria.10.1109/SENSORS56945.2023.10324957 ↗
10. Kang, S.w., Jang, M.h., and Lee, S. (2021). Identification of Human Motion Using Radar Sensor in an Indoor Environment. Sensors, 21.10.3390/s21072305 ↗
11. Zhao, Y., Zhou, H., Lu, S., Liu, Y., An, X., and Liu, Q. (2022). Human Activity Recognition Based on Non-Contact Radar Data and Improved PCA Method. Appl. Sci., 12.10.3390/app12147124 ↗
12. HRSpecNET: A Deep Learning-Based High-Resolution Radar Micro-Doppler Signature Reconstruction for Improved HAR Classification10.1109/TRS.2024.3396172 ↗
13. Tan, T.H., Tian, J.H., Sharma, A.K., Liu, S.H., and Huang, Y.F. (2024). Human Activity Recognition Based on Deep Learning and Micro-Doppler Radar Data. Sensors, 24.10.3390/s24082530 ↗
14. Cao, L., Liang, S., Zhao, Z., Wang, D., Fu, C., and Du, K. (2023). Human Activity Recognition Method Based on FMCW Radar Sensor with Multi-Domain Feature Attention Fusion Network. Sensors, 23.10.3390/s23115100 ↗
15. Li, Z., Fioranelli, F., Yang, S., Zhang, L., Romain, O., He, Q., Cui, G., and Le Kernec, J. (2020, January 4–6). Multi-domains based human activity classification in radar. Proceedings of the IET International Radar Conference (IET IRC 2020), Virtual.10.1049/icp.2021.0557 ↗
16. Huang, L., Lei, D., Zheng, B., Chen, G., An, H., and Li, M. (2024). Lightweight Multi-Domain Fusion Model for Through-Wall Human Activity Recognition Using IR-UWB Radar. Appl. Sci., 14.10.3390/app14209522 ↗
17. Radar-Based Human-Motion Recognition With Deep Learning: Promising Applications for Indoor Monitoring10.1109/MSP.2018.2890128 ↗
18. Yang, X., Guendel, R.G., Yarovoy, A., and Fioranelli, F. (2022, January 21–25). Radar-based Human Activities Classification with Complex-valued Neural Networks. Proceedings of the 2022 IEEE Radar Conference (RadarConf22), New York, NY, USA.10.1109/RadarConf2248738.2022.9763903 ↗
19. Li, H., Shrestha, A., Heidari, H., Kernec, J.L., and Fioranelli, F. (2019, January 6–8). Activities Recognition and Fall Detection in Continuous Data Streams Using Radar Sensor. Proceedings of the 2019 IEEE MTT-S International Microwave Biomedical Conference (IMBioC), Nanjing, China.10.1109/IMBIOC.2019.8777855 ↗
20. Continuous Human Activity Classification From FMCW Radar With Bi-LSTM Networks10.1109/JSEN.2020.3006386 ↗
21. Continuous Human Activity Classification With Unscented Kalman Filter Tracking Using FMCW Radar10.1109/LSENS.2020.2991367 ↗
22. Guendel, R.G., Fioranelli, F., and Yarovoy, A. (2020, January 21–25). Derivative Target Line (DTL) for Continuous Human Activity Detection and Recognition. Proceedings of the 2020 IEEE Radar Conference (RadarConf20), Florence, Italy.10.1109/RadarConf2043947.2020.9266383 ↗
23. Kurtoglu, E., Gurbuz, A.C., Malaia, E., Griffin, D., Crawford, C., and Gurbuz, S.Z. (2021, January 8–14). Sequential Classification of ASL Signs in the Context of Daily Living Using RF Sensing. Proceedings of the 2021 IEEE Radar Conference (RadarConf21), Atlanta, GA, USA.10.1109/RadarConf2147009.2021.9455178 ↗
24. Continuous Human Activity Recognition with Distributed Radar Sensor Networks and CNN–RNN Architectures10.1109/TGRS.2022.3230829 ↗
25. Distributed radar fusion and recurrent networks for classification of continuous human activities10.1049/rsn2.12249 ↗
26. Qu, L., Li, X., Yang, T., and Wang, S. (2025). Radar-Based Continuous Human Activity Recognition Using Multidomain Fusion Vision Transformer. IEEE Sens. J.10.1109/JSEN.2025.3535673 ↗
27. Feng, X., Chen, P., Weng, Y., and Zheng, H. (2025). CMDN: Continuous Human Activity Recognition Based on Multi-domain Radar Data Fusion. IEEE Sens. J.10.1109/JSEN.2025.3530921 ↗
28. Classification Strategies for Radar-Based Continuous Human Activity Recognition With Multiple Inputs and Multilabel Output10.1109/JSEN.2024.3429549 ↗
29. ASL Trigger Recognition in Mixed Activity/Signing Sequences for RF Sensor-Based User Interfaces10.1109/THMS.2021.3131675 ↗
30. Kruse, N., Guendel, R., Fioranelli, F., and Yarovoy, A. (2022, January 24–27). Segmentation of Micro-Doppler Signatures of Human Sequential Activities using Rényi Entropy. Proceedings of the International Conference on Radar Systems (RADAR 2022), Edinburgh, UK.10.1049/icp.2022.2357 ↗
31. Liu, S., and Wang, B. (2024, January 24–26). Optimized Modified ResNet18: A Residual Neural Network for High Resolution. Proceedings of the 2024 IEEE 4th International Conference on Electronic Technology, Communication and Information (ICETCI), Changchun, China.10.1109/ICETCI61221.2024.10594672 ↗
32. Guendel, R.G., Unterhorst, M., Fioranelli, F., and Yarovoy, A. (2021). Dataset of continuous human activities performed in arbitrary directions collected with a distributed radar network of five nodes. 4TU. ResearchData, 10.10.1109/RadarConf2147009.2021.9454972 ↗