Hybrid Deep Learning for Human Fall Detection: A Synergistic Approach Using YOLOv8 and Time-Space Transformers
Falls are a major concern, particularly for elderly individuals and vulnerable populations, often leading to severe injuries if not detected promptly. Traditional sensor-based fall detection methods suffer from limitations such as discomfort, maintenance challenges, and susceptibility to false readi...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10909442/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850069619703808000 |
|---|---|
| author | Yousef Sanjalawe Salam Fraihat Mosleh Abualhaj Salam R. Al-E'Mari Emran Alzubi |
| author_facet | Yousef Sanjalawe Salam Fraihat Mosleh Abualhaj Salam R. Al-E'Mari Emran Alzubi |
| author_sort | Yousef Sanjalawe |
| collection | DOAJ |
| description | Falls are a major concern, particularly for elderly individuals and vulnerable populations, often leading to severe injuries if not detected promptly. Traditional sensor-based fall detection methods suffer from limitations such as discomfort, maintenance challenges, and susceptibility to false readings. To address these issues, we propose a hybrid deep learning-based system that leverages video-based techniques for accurate and efficient fall detection. Our approach integrates YOLOv8 for real-time human detection with Time-Space Transformers for temporal motion analysis, ensuring robust recognition of fall incidents while minimizing false positives and negatives. The proposed system is evaluated on two benchmark datasets, CUCAFall and DiverseFALL10500, which contain diverse and challenging environmental scenarios. Experimental results demonstrate that our model outperforms existing state-of-the-art methods, achieving a mean Average Precision (mAP) of 0.9955 on CUCAFall and 0.950 on DiverseFALL10500, along with F1-scores exceeding 0.998 for all classes in certain configurations. These results confirm the model’s ability to distinguish fall events from normal activities accurately. Furthermore, the system maintains computational efficiency, making it suitable for real-time deployment in healthcare and smart home environments. By enhancing spatial and temporal awareness, our hybrid approach significantly improves fall detection reliability, ensuring timely intervention and enhancing safety for at-risk individuals. |
| format | Article |
| id | doaj-art-67aa71107daf44bd95f7167f3e68872d |
| institution | DOAJ |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-67aa71107daf44bd95f7167f3e68872d2025-08-20T02:47:44ZengIEEEIEEE Access2169-35362025-01-0113413364136610.1109/ACCESS.2025.354791410909442Hybrid Deep Learning for Human Fall Detection: A Synergistic Approach Using YOLOv8 and Time-Space TransformersYousef Sanjalawe0https://orcid.org/0000-0002-4442-1865Salam Fraihat1https://orcid.org/0000-0002-1025-7868Mosleh Abualhaj2https://orcid.org/0000-0002-7465-8038Salam R. Al-E'Mari3https://orcid.org/0000-0002-2134-4158Emran Alzubi4Department of Information Technology, King Abdullah II School for Information Technology, The University of Jordan (JU), Amman, JordanArtificial Intelligence Research Center (AIRC), College of Engineering and Information Technology, Ajman University, Ajman, United Arab EmiratesDepartment of Networks and Information Security, Faculty of Information Technology, Al-Ahliyya Amman University, Amman, JordanInformation Security Department, Faculty of Information Technology, University of Petra, Amman, JordanCollege of Business Administration, Northern Border University (NBU), Arar, Saudi ArabiaFalls are a major concern, particularly for elderly individuals and vulnerable populations, often leading to severe injuries if not detected promptly. Traditional sensor-based fall detection methods suffer from limitations such as discomfort, maintenance challenges, and susceptibility to false readings. To address these issues, we propose a hybrid deep learning-based system that leverages video-based techniques for accurate and efficient fall detection. Our approach integrates YOLOv8 for real-time human detection with Time-Space Transformers for temporal motion analysis, ensuring robust recognition of fall incidents while minimizing false positives and negatives. The proposed system is evaluated on two benchmark datasets, CUCAFall and DiverseFALL10500, which contain diverse and challenging environmental scenarios. Experimental results demonstrate that our model outperforms existing state-of-the-art methods, achieving a mean Average Precision (mAP) of 0.9955 on CUCAFall and 0.950 on DiverseFALL10500, along with F1-scores exceeding 0.998 for all classes in certain configurations. These results confirm the model’s ability to distinguish fall events from normal activities accurately. Furthermore, the system maintains computational efficiency, making it suitable for real-time deployment in healthcare and smart home environments. By enhancing spatial and temporal awareness, our hybrid approach significantly improves fall detection reliability, ensuring timely intervention and enhancing safety for at-risk individuals.https://ieeexplore.ieee.org/document/10909442/Deep fall detectionCUCAFallDiverseFALL10500human detectiontime-space transformersYOLOv8 |
| spellingShingle | Yousef Sanjalawe Salam Fraihat Mosleh Abualhaj Salam R. Al-E'Mari Emran Alzubi Hybrid Deep Learning for Human Fall Detection: A Synergistic Approach Using YOLOv8 and Time-Space Transformers IEEE Access Deep fall detection CUCAFall DiverseFALL10500 human detection time-space transformers YOLOv8 |
| title | Hybrid Deep Learning for Human Fall Detection: A Synergistic Approach Using YOLOv8 and Time-Space Transformers |
| title_full | Hybrid Deep Learning for Human Fall Detection: A Synergistic Approach Using YOLOv8 and Time-Space Transformers |
| title_fullStr | Hybrid Deep Learning for Human Fall Detection: A Synergistic Approach Using YOLOv8 and Time-Space Transformers |
| title_full_unstemmed | Hybrid Deep Learning for Human Fall Detection: A Synergistic Approach Using YOLOv8 and Time-Space Transformers |
| title_short | Hybrid Deep Learning for Human Fall Detection: A Synergistic Approach Using YOLOv8 and Time-Space Transformers |
| title_sort | hybrid deep learning for human fall detection a synergistic approach using yolov8 and time space transformers |
| topic | Deep fall detection CUCAFall DiverseFALL10500 human detection time-space transformers YOLOv8 |
| url | https://ieeexplore.ieee.org/document/10909442/ |
| work_keys_str_mv | AT yousefsanjalawe hybriddeeplearningforhumanfalldetectionasynergisticapproachusingyolov8andtimespacetransformers AT salamfraihat hybriddeeplearningforhumanfalldetectionasynergisticapproachusingyolov8andtimespacetransformers AT moslehabualhaj hybriddeeplearningforhumanfalldetectionasynergisticapproachusingyolov8andtimespacetransformers AT salamralemari hybriddeeplearningforhumanfalldetectionasynergisticapproachusingyolov8andtimespacetransformers AT emranalzubi hybriddeeplearningforhumanfalldetectionasynergisticapproachusingyolov8andtimespacetransformers |