Enhancing cross-dataset performance in distracted driver detection using body part activity recognition.
| dc.contributor.advisor | Nicolls, Frederick | |
| dc.contributor.advisor | Stoltz, Gene | |
| dc.contributor.author | Zandamela, Frank | |
| dc.date.accessioned | 2025-03-12T09:05:39Z | |
| dc.date.available | 2025-03-12T09:05:39Z | |
| dc.date.issued | 2024-05 | |
| dc.date.updated | 2025-03-12T08:55:32Z | |
| dc.description.abstract | Detecting distracted drivers is a crucial task, and the literature proposes various deep learning-based methods. Among these methods, convolutional neural networks dominate because they can extract and learn image features automatically. However, even though existing methods have reported remarkable results, the cross-dataset performance of these methods remains unknown. A problem arises because cross-dataset performance often indicates a model's generalisation ability. Without knowing the model's cross-dataset performance, deployment in the real world could result in catastrophic events. This thesis investigates the generalisation ability of deep learning-based distracted driver detection methods. In addition, a robust distracted driver detection approach is proposed. The proposed approach is based on recognising distinctive activities of human body parts involved when a driver is operating a vehicle. Representative state-of-the-art deep learning-based methods have been trained exclusively on three widely used image datasets and evaluated across the test sets of these datasets. Experimental results reveal that current deep learning-based methods for detecting distracted drivers do not generalise well on unknown datasets, particularly for convolutional neural network (CNN) models that use the entire image for prediction. In addition, the experiments indicated that although current distracted driver detection datasets are relatively large, they lack diversity. The proposed approach was implemented using a state-of-the-art object detection algorithm called Yolov7. The cross-dataset performance of the implemented approach was evaluated on three benchmark datasets and a custom dataset. Experimental results demonstrate that the proposed approach improves cross-dataset performance. A cross-dataset accuracy improvement of 7.8% was observed. Most importantly, the overall balanced (F1-score) performance was improved by a factor of 2.68. The experimental results also revealed that although the proposed approach demonstrates commendable performance on a custom test set, all algorithms encountered challenges when dealing with the custom test set, mainly due to lower image quality and difficult lighting conditions. The thesis presents two main contributions. Firstly, it evaluates the performance of current deep learning-based distracted driver detection algorithms across different datasets. Secondly, it proposes a robust algorithm for detecting distracted drivers by identifying key human body parts involved in operating a vehicle. | |
| dc.identifier.apacitation | Zandamela, F. (2024). <i>Enhancing cross-dataset performance in distracted driver detection using body part activity recognition</i>. (). University of Cape Town ,Faculty of Engineering and the Built Environment ,Department of Electrical Engineering. Retrieved from http://hdl.handle.net/11427/41155 | en_ZA |
| dc.identifier.chicagocitation | Zandamela, Frank. <i>"Enhancing cross-dataset performance in distracted driver detection using body part activity recognition."</i> ., University of Cape Town ,Faculty of Engineering and the Built Environment ,Department of Electrical Engineering, 2024. http://hdl.handle.net/11427/41155 | en_ZA |
| dc.identifier.citation | Zandamela, F. 2024. Enhancing cross-dataset performance in distracted driver detection using body part activity recognition. . University of Cape Town ,Faculty of Engineering and the Built Environment ,Department of Electrical Engineering. http://hdl.handle.net/11427/41155 | en_ZA |
| dc.identifier.ris | TY - Thesis / Dissertation AU - Zandamela, Frank AB - Detecting distracted drivers is a crucial task, and the literature proposes various deep learning-based methods. Among these methods, convolutional neural networks dominate because they can extract and learn image features automatically. However, even though existing methods have reported remarkable results, the cross-dataset performance of these methods remains unknown. A problem arises because cross-dataset performance often indicates a model's generalisation ability. Without knowing the model's cross-dataset performance, deployment in the real world could result in catastrophic events. This thesis investigates the generalisation ability of deep learning-based distracted driver detection methods. In addition, a robust distracted driver detection approach is proposed. The proposed approach is based on recognising distinctive activities of human body parts involved when a driver is operating a vehicle. Representative state-of-the-art deep learning-based methods have been trained exclusively on three widely used image datasets and evaluated across the test sets of these datasets. Experimental results reveal that current deep learning-based methods for detecting distracted drivers do not generalise well on unknown datasets, particularly for convolutional neural network (CNN) models that use the entire image for prediction. In addition, the experiments indicated that although current distracted driver detection datasets are relatively large, they lack diversity. The proposed approach was implemented using a state-of-the-art object detection algorithm called Yolov7. The cross-dataset performance of the implemented approach was evaluated on three benchmark datasets and a custom dataset. Experimental results demonstrate that the proposed approach improves cross-dataset performance. A cross-dataset accuracy improvement of 7.8% was observed. Most importantly, the overall balanced (F1-score) performance was improved by a factor of 2.68. The experimental results also revealed that although the proposed approach demonstrates commendable performance on a custom test set, all algorithms encountered challenges when dealing with the custom test set, mainly due to lower image quality and difficult lighting conditions. The thesis presents two main contributions. Firstly, it evaluates the performance of current deep learning-based distracted driver detection algorithms across different datasets. Secondly, it proposes a robust algorithm for detecting distracted drivers by identifying key human body parts involved in operating a vehicle. DA - 2024-05 DB - OpenUCT DP - University of Cape Town KW - Engineering LK - https://open.uct.ac.za PB - University of Cape Town PY - 2024 T1 - Enhancing cross-dataset performance in distracted driver detection using body part activity recognition TI - Enhancing cross-dataset performance in distracted driver detection using body part activity recognition UR - http://hdl.handle.net/11427/41155 ER - | en_ZA |
| dc.identifier.uri | http://hdl.handle.net/11427/41155 | |
| dc.identifier.vancouvercitation | Zandamela F. Enhancing cross-dataset performance in distracted driver detection using body part activity recognition. []. University of Cape Town ,Faculty of Engineering and the Built Environment ,Department of Electrical Engineering, 2024 [cited yyyy month dd]. Available from: http://hdl.handle.net/11427/41155 | en_ZA |
| dc.language.iso | en | |
| dc.language.rfc3066 | Eng | |
| dc.publisher.department | Department of Electrical Engineering | |
| dc.publisher.faculty | Faculty of Engineering and the Built Environment | |
| dc.publisher.institution | University of Cape Town | |
| dc.subject | Engineering | |
| dc.title | Enhancing cross-dataset performance in distracted driver detection using body part activity recognition. | |
| dc.type | Thesis / Dissertation | |
| dc.type.qualificationlevel | Masters | |
| dc.type.qualificationlevel | MSc |