Convxpf-Net: An Explainable Ai Approach For Recognizing Pedestrian Behaviour Using Convolutional Xgboost Pyramid Fractional Network

Dr. Amol Raoji Madane; Dr. Ho Chin Kuan

doi:10.29284/1aeeq569

Authors

Dr. Amol Raoji Madane Post-Doc fellow, Asia Pacific University of Technology, Jalan Technologi 5, Taman Technologi Malaysia
Dr. Ho Chin Kuan Vice Chancellor and Post-Doc Research Guide, Asia Pacific University of Technology, Jalan Technologi 5, Taman Technologi Malaysia

DOI:

https://doi.org/10.29284/1aeeq569

Keywords:

Pedestrian Behaviour, Advanced Driver Assistance System, Deep Learning, Pedestrian Detection and Classification.

Abstract

In Advanced Driver Assistance System (ADAS), identifying pedestrians and predicting their future behaviour is essential for preventing accidents, however during heavy occlusion, the prevailing Pedestrian Behaviour Detection (PBD) techniques fails to identify the appropriate behaviour. Therefore, in this research, a novel Explainable AI based Convolutional XGBoost pyramid fractional Network (ConvXPF-Net) is proposed. The existing PBD techniques struggles to identify the patterns in the frames due to orientation-invariant regression head, and hence, YOLOv8 based Visual Casual Model (Y8-VCM) is employed, which eliminates the Boundary Discontinuity problem in Angle prediction. Besides, for addressing missed detections and issues with distracted walking during classification, a Convolutional XGBoost Pyramid Fractional Network (ConvXPF-Net) is utilized. The complexities in detecting the pedestrians with less radial velocities due to minimal Doppler shift, is reduced by Pyramid Net Technique (PNT), and then for recognizing the subtle behavioural changes that leads to oversimplification of the features, a Fractional Calculus Technique is employed. Further, based on the classified PB, the movement of pedestrians is determined, which further aids in performing speed and brake automation. Finally, the proposed work is validated using Explainable AI, and the outcomes attained shows the effectiveness of proposed work in Pedestrian behaviour prediction.

References

[1]. Zahedi, H., Mohammadinia, L., Dehghani, S.L., Habibzadeh, S. and Kheibar, N., 2024. The association between health literacy and pedestrian safety behavior among adults: a cross-sectional study. BMC Public Health, 24(1), p.1110.

[2]. Acela, L.M.B., Sousa, E., Faria, S., Almeida, R. and Freitas, E.F., 2025. Pedestrian crossing behavior in social groups: Exploring cultural contexts through a comparative study. Case Studies on Transport Policy, 19, p.101333.

[3]. Cai, J., Wang, M. and Wu, Y., 2024. Research on pedestrian crossing decision models and predictions based on machine learning. Sensors, 24(1), p.258.

[4]. de Winkel, K.N. and Christoph, M., 2025. Rethinking Advanced Driver Assistance System taxonomies: A framework and inventory of real-world safety performance. Transportation Research Interdisciplinary Perspectives, 29, p.101336.

[5]. Hussien, M.M., Melo, A.N., Ballardini, A.L., Maldonado, C.S., Izquierdo, R. and Sotelo, M.A., 2025. Rag-based explainable prediction of road users behaviors for automated driving using knowledge graphs and large language models. Expert Systems with Applications, 265, p.125914.

[6]. Guo, Y., Ma, S., Wei, F., Lu, L., Sun, F. and Wang, J., 2022. Analysis of behavior characteristics for pedestrian twice-crossing at signalized intersections based on an improved social force model. Sustainability, 14(4), p.2003.

[7]. Reyes-Norambuena, P., Pinto, A.A., Martínez, J., Karbassi Yazdi, A. and Tan, Y., 2024. The Application of Machine Learning and Deep Learning with a Multi-Criteria Decision Analysis for Pedestrian Modeling: A Systematic Literature Review (1999–2023). Sustainability, 17(1), p.41.

[8]. Liang, E. and Stamp, M., 2023. Predicting pedestrian crosswalk behavior using Convolutional Neural Networks. Traffic injury prevention, 24(4), pp.338-343.

[9]. Alghodhaifi, H. and Lakshmanan, S., 2023. Holistic Spatio-Temporal graph attention for trajectory prediction in Vehicle–Pedestrian interactions. Sensors, 23(17), p.7361.

[10]. Ham, J.S., Kim, D.H. and Moon, J., 2025. WatchoutPed: A dataset and model for Vulnerable Pedestrian Anticipation in surveillance videos. Knowledge-Based Systems, p.113352.

[11]. Iaboni, C., Kelly, T. and Abichandani, P., 2025. NU-AIR: A Neuromorphic Urban Aerial Dataset for Detection and Localization of Pedestrians and Vehicles. International Journal of Computer Vision, pp.1-19.

[12]. Liu, S., Zhang, L., Ge, J., Li, W. and Long, Y., 2025. Framework for timely perception of spatiotemporal crowd congestion risk in public spaces based on video pedestrian tracking and geographic mapping. GIScience & Remote Sensing, 62(1), p.2480416.

[13]. Nomoto, K. and Yifan, C., 2025. Analysis of Increase in Accidents Between Pedestrians and Vehicles During Twilight: A Perspective of Pedestrians. IEEE Access.

[14]. Pipkorn, L., Domeyer, J., Mehler, B., Reimer, B. and Gershon, P., 2025. Decoding the silent dialogue: Unveiling driver-pedestrian communication dynamics with a hidden Markov model. Transportation Research Part F: Traffic Psychology and Behaviour, 109, pp.965-976.

[15]. Ahsan, M.J., Abdel-Aty, M., Anwari, N. and Barbour, N., 2025. Evaluating drivers’ braking behavior at mid-block pedestrian crosswalks using video data and a mixed logit model with heterogeneity in the means. Transportation Research Interdisciplinary Perspectives, 31, p.101380.

[16]. Hassani, S., Mustapha, S., Li, J., Mousavi, M. and Dackermann, U., 2025. Next-generation coupled structure-human sensing technology: Enhanced pedestrian-bridge interaction analysis using data fusion and machine learning. Information Fusion, p.102983.

[17]. Han, C., Seshadri, P., Ding, Y., Posner, N., Koo, B.W., Agrawal, A., Lerch, A. and Guhathakurta, S., 2024. Understanding pedestrian movement using urban sensing technologies: the promise of audio-based sensors. Urban Informatics, 3(1), p.22.

[18]. BaniSalman, M., Aljaidi, M., Elgeberi, N., Alsarhan, A. and Al Mamlook, R.E., 2025. VRDeepSafety: A Scalable VR Simulation Platform with V2X Communication for Enhanced Accident Prediction in Autonomous Vehicles. World Electric Vehicle Journal, 16(2), p.82.

[19]. Iftikhar, S., Asim, M., Zhang, Z. and El-Latif, A.A.A., 2022. Advance generalization technique through 3D CNN to overcome the false positives pedestrian in autonomous vehicles. Telecommunication Systems, 80(4), pp.545-557.

[20]. Yang, E., Shankar, K., Kumar, S. and Seo, C., 2023. Bioinspired Garra Rufa optimization-assisted deep learning model for object classification on pedestrian walkways. Biomimetics, 8(7), p.541.

[21]. Aung, N.H.H., Sangwongngam, P., Jintamethasawat, R. and Wuttisittikulkij, L., 2024. Fully Synthetic Pedestrian Anomaly Behavior Dataset Generation in Metaverse for Enhancing Autonomous Driving Object Detection. IEEE Access.

[22]. Kilicarslan, M. and Zheng, J.Y., 2022. DeepStep: Direct detection of walking pedestrian from motion by a vehicle camera. IEEE Transactions on Intelligent Vehicles, 8(2), pp.1652-1663.

[23]. Siva Senthil, D. and Sivarani, T.S., 2025. SED-NET: Real-Time Suspicious Event Detection via Deep Learning-Based Di-Stream Neural Network. International Journal of Computational Intelligence Systems, 18(1), pp.1-19.

[24]. Xu, J., Xu, D., Wu, J. and Shi, X., 2025. Modeling the collective behavior of pedestrians with the spontaneous loose leader–follower structure in public spaces. Computer‐Aided Civil and Infrastructure Engineering.

[25]. Jain, D.K., Zhao, X., Garcia, S. and Neelakandan, S., 2024. Robust multi-modal pedestrian detection using deep convolutional neural network with ensemble learning model. Expert Systems with Applications, 249, p.123527.

[26]. Kaur, S., Kaur, L. and Lal, M., 2024. An effective Key Frame Extraction technique based on Feature Fusion and Fuzzy-C means clustering with Artificial Hummingbird. Scientific Reports, 14(1), p.26651.

[27]. Yaseen, M., 2024. What is yolov9: An in-depth exploration of the internal features of the next-generation object detector. arXiv preprint arXiv:2409.07813.

[28]. Liu, L., Sun, S., Zhi, S., Shi, F., Liu, Z., Heikkilä, J. and Liu, Y., 2025. A Causal Adjustment Module for Debiasing Scene Graph Generation. IEEE Transactions on Pattern Analysis and Machine Intelligence.

[29]. Shrivastava, A. and Poonkuntran, S.H.A.N.M.U.G.A.M., 2025. Multi-Scale Feature Pyramid Network with Camera Artifact Augmentation for Pedestrian Detection. IEEE Access.

[30]. Coelho, C., Costa, M.F.P. and Ferrás, L.L., 2024. Fractional Calculus meets neural networks for Computer Vision: a survey. AI, 5(3), pp.1391-1426.

[31]. Galli, A., La Gatta, V., Moscato, V., Postiglione, M. and Sperlì, G., 2024. Explainability in AI-based behavioral malware detection systems. Computers & Security, 141, p.103842.

[32]. Panduru, K. and Walsh, J., 2025. Exploring the Unseen: A Survey of Multi-Sensor Fusion and the Role of Explainable AI (XAI) in Autonomous Vehicles. Sensors (Basel, Switzerland), 25(3), p.856.

[33]. Mansouri, W., Alohali, M.A., Alqahtani, H., Alruwais, N., Alshammeri, M. and Mahmud, A., 2025. Deep convolutional neural network-based enhanced crowd density monitoring for intelligent urban planning on smart cities. Scientific Reports, 15(1), p.5759.

[34]. Su, J., Wang, F. and Zhuang, W., 2025. An Improved YOLOv7 Tiny Algorithm for Vehicle and Pedestrian Detection with Occlusion in Autonomous Driving. Chinese Journal of Electronics, 34(1), pp.282-294.

[35]. Zhang, H., Deng, J., Xu, Y., Deng, Y. and Lin, J.R., 2025. An Adaptive Pedestrian Flow Prediction Model Based on First-Order Differential Error Adjustment and Hidden Markov Model. Buildings, 15(6), p.902.

[36]. Yan, S., Luo, B., Sun, X., Xiao, J., Ji, Y. and Ghazali, K.H.B., 2025. Accurate Localization Method Combining Optimized Hybrid Neural Networks for Geomagnetic Localization with Multi-Feature Dead Reckoning. Sensors, 25(5), p.1304.

[37]. Zuo, Y., Jiang, J. and Yada, K., 2025. Application of hybrid gate recurrent unit for in-store trajectory prediction based on indoor location system. Scientific Reports, 15(1), p.1055.

[38]. Sivalingan, H., 2024. Pedestrian Detection In Video Surveillance Using Yolo V5 With Light Perception Fusion. ICTACT Journal on Soft Computing, 14(4).

[39]. Tahir, M. and Anwar, S., 2021. Transformers in pedestrian image retrieval and person re-identification in a multi-camera surveillance system. Applied Sciences, 11(19), p.9197.

[40]. Haggui, O., Bayd, H. and Magnier, B., 2024. Centroid human tracking via oriented detection in overhead fisheye sequences. The Visual Computer, 40(1), pp.407-425.

[41]. https://data.nvision2.eecs.yorku.ca/JAAD_dataset/

[42]. Astha Modak, Samruddhi Paradkar, Shruti Manwatkar, Amol R. Madane, and Ashwini M. Deshpande. "Human head pose and eye state-based driver distraction monitoring system." In Proceedings of 3rd International Conference on Computer Vision and Image Processing: CVIP 2018, Volume 1, pp. 403-421. Singapore: Springer Singapore, 2019.

[43]. Dinesh Kadam, Amol R. Madane, Krishnan Kutty, S. V. Bonde, “Rain Streaks Elimination using Image Processing Algorithms”, Signal & Image Processing: An International Journal (SIPIJ), Vol. 10, No. 3, pp. 21-32, Jun-2019

[44]. Apeksha Kamble, Akanksha Singh, Damini Thusoo, Amol R. Madane, "Design and Implementation of Surround View Monitoring System in View of Autonomous Vehicle", 6th International Conference on Data Management, Analytics and Innovation 2022 (ICDMAI 2022), India, pp. 509-519, 14-16 Jan 2022.

[45]. Ashish Kathe, Pranav Teli, Amol R. Madane, “COVID-19 Regulations Check: Social Distancing, People Counting and Mask Wear Check”, 6th International Conference on Data Management, Analytics and Innovation 2022 (ICDMAI 2022), India, pp. 467-484, 14-16 Jan 2022.

[46]. Priyojit Chakraborty, Amol R. Madane, “Vehicle Queue Management System at Fuel Station”, NIELIT's International Conference on Communication, Electronics and Digital Technologies (NICE-DT'23), New Delhi, India, 10-11 Feb 2023.ss