Abstract
Video Anomaly detection has been the focus of many research studies for a long time and shows potential for endless implementations to detect real-world anomalies. When it comes to road safety, Jaywalking is one such real-life event that can be detected and localized using video anomaly detection techniques. However, this field’s progress depends majorly on the availability of diverse datasets and the kind of anomalies that they depict. The main objective of this work is to propose a video anomaly detection model for jaywalking that can help control the number of road accidents that occur due to jaywalking each year. The proposed model is a novel variation of the InceptionV3 deep CNN model and has been experimenting with one of the latest street scene datasets and its different variations. Our model consists of two separate subsystems based on pre-trained InceptionV3 architecture. The first subsystem, Anomaly Detector, takes a video frame as input and predicts whether it consists of a jaywalking event. The second subsystem, Anomaly Localizer, takes a jaywalking(anomalous) video frame as input and predicts the bounding box labels to locate the Jaywalking object in the frame. To evaluate, we employ the Street Scene Dataset, released in February 2020 that offers several kinds of jaywalking events. In this work, we experiment with different pre-trained CNN models, namely VGG16, ResNet50, and InceptionV3, to evaluate their performance for anomaly detection. Further, we assess the model’s performance for different dataset sizes, having both an even and an uneven number of anomalous and non-anomalous frames. Lastly, we measure the effectiveness of the model on different types of jaywalking episodes. The evaluation shows that our proposed model attains remarkable detection accuracy, in comparison to other state-of-the-art methods proposed previously in this field.
Similar content being viewed by others
References
Adam A, Rivlin E, Shimshoni I, Reinitz D (2008) Robust real-time unusual event detection using multiple fixed-location monitors. IEEE transactions on pattern analysis and machine intelligence 30(3):555–560
Bansod S, Nandedkar A (2019) Transfer learning for video anomaly detection. Journal of Intelligent & Fuzzy Systems 36(3):1967–1975
Calderara S, Heinemann U, Prati A, Cucchiara R, Tishby N (2011) Detecting anomalies in people’s trajectories using spectral graph analysis. Comput Vis Image Underst 115(8):1099–1111
Highways of Road Transport M (2019) Road accidents in india – 2019. Tech. rep., Government of India, https://morth.nic.in/road-accident-in-india
Jiang X, Wang W, Mao Y, Bengler K, Bubb H (2011) Situational factors of influencing drivers to give precedence to jaywalking pedestrians at signalized crosswalk. International Journal of Computational Intelligence Systems 4(6):1407–1414
John V, Boyali A, Thompson S, Lakshmanan A, Mita S (2020) Visible and thermal camera-based jaywalking estimation using a hierarchical deep learning framework. In: Proceedings of the Asian Conference on Computer Vision
Kaltsa V, Briassouli A, Kompatsiaris I, Strintzis M G (2018) Multiple hierarchical dirichlet processes for anomaly detection in traffic. Comput Vis Image Underst 169:28–39
Li W, Mahadevan V, Vasconcelos N (2013) Anomaly detection and localization in crowded scenes. IEEE transactions on pattern analysis and machine intelligence 36(1):18–32
Liu W, Luo W, Lian D, Gao S (2018) Future frame prediction for anomaly detection–a new baseline. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 6536–6545
Loy C C, Xiang T, Gong S (2008) From local temporal correlation to global anomaly detection. In: The 1st International Workshop on Machine Learning for Vision-based Motion Analysis-MLVMA’08
Lu C, Shi J, Jia J (2013) Abnormal event detection at 150 fps in matlab. In: Proceedings of the IEEE international conference on computer vision, pp 2720–2727
Luo W, Liu W, Gao S (2017) A revisit of sparse coding based anomaly detection in stacked rnn framework. In: Proceedings of the IEEE International Conference on Computer Vision, pp 341–349
Medel J R, Savakis A (2016) Anomaly detection in video using predictive convolutional long short-term memory networks. arXiv:161200390
Morris B T, Trivedi M M (2008) A survey of vision-based trajectory learning and analysis for surveillance. IEEE transactions on circuits and systems for video technology 18(8):1114–1127
Park J, Lee Y, Heo J H, Kang S J (2019) Convolutional neural network-based jaywalking data generation and classification. In: 2019 International soc design conference, ISOCC, IEEE, pp 132–133
Ramachandra B, Jones M (2020) Street scene: a new dataset and evaluation protocol for video anomaly detection. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pp 2569–2578
Sabokrou M, Fathy M, Hoseini M (2016) Video anomaly detection and localisation based on the sparsity and reconstruction error of auto-encoder. Electron Lett 52(13):1122–1124
Sabokrou M, Fayyaz M, Fathy M, Klette R (2017) Deep-cascade: cascading 3d deep neural networks for fast anomaly detection and localization in crowded scenes. IEEE Trans Image Process 26(4):1992–2004
Sabokrou M, Fayyaz M, Fathy M, Moayed Z, Klette R (2018) Deep-anomaly: fully convolutional neural network for fast anomaly detection in crowded scenes. Comput Vis Image Underst 172:88–97
Sayed T, Zaki M H, Autey J (2013) A Novel approach for diagnosing road safety issues using automated computer vision techniques. In: 16th International Conference Road Safety on Four Continents. Beijing, China (RS4c 2013), pp 15–17, May 2013, Statens väg-ochtransportforskningsinstitut
Sivaraman S, Trivedi M M (2013) Looking at vehicles on the road: a survey of vision-based vehicle detection, tracking, and behavior analysis. IEEE transactions on intelligent transportation systems 14(4):1773–1795
Suarez J J P, Naval Jr P C (2020) A survey on deep learning techniques for video anomaly detection. arXiv:200914146
Sultani W, Chen C, Shah M (2018) Real-world anomaly detection in surveillance videos. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 6479–6488
Szegedy C, Vanhoucke V, Ioffe S, Shlens J, Wojna Z (2016) Rethinking the inception architecture for computer vision. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 2818–2826
Tian B, Yao Q, Gu Y, Wang K, Li Y (2011) Video Processing techniques for traffic flow monitoring: A survey. In: 2011 14Th international IEEE conference on intelligent transportation systems, ITSC, IEEE, pp 1103–1108
Tung F, Zelek J S, Clausi D A (2011) Goal-based trajectory analysis for unusual behaviour detection in intelligent surveillance. Image Vis Comput 29(4):230–240
Vishwakarma S, Agrawal A (2013) A survey on activity recognition and behavior understanding in video surveillance. Vis Comput 29(10):983–1009
WHO (2020) Road traffic injuries. Tech. rep., https://www.who.int/news-room/fact-sheets/detail/road-traffic-injuries
Ye M, Peng X, Gan W, Wu W, Qiao Y (2019) Anopcn: video anomaly detection via deep predictive coding network. In: Proceedings of the 27th ACM International Conference on Multimedia, pp 1805–1813
Yu J, Li X, Zhao Y (2011) Jaywalker detection based on motion path analysis. In: Third International Conference on Transportation Engineering (ICTE) American Society of Civil EngineersChina Communications and Transportation Association
Zaki M H, Sayed T, Tageldin A, Hussein M (2013) Application of computer vision to diagnosis of pedestrian safety issues. Transportation research record 2393(1):75–84
Zhang Y, Lu H, Zhang L, Ruan X (2016) Combining motion and appearance cues for anomaly detection. Pattern Recogn 51:443–452
Acknowledgements
We acknowledge the authors of our base paper [16] to release the dataset comprising Jaywalking that was focus of our work.
Funding
There is no funding source of the work and project.
Author information
Authors and Affiliations
Contributions
Aarti Bala performed implementation of models, analyzed the results, and wrote the first verion of this draft. Rishabh Kaushal conceptualization the problem problem, reviewed the methodology, code, results, and write-up in this report.
Corresponding author
Ethics declarations
This is to declare that we shall release the code and customized dataset that we curated from the original dataset on acceptance of our work.
Conflict of Interests
There are no conflicts of interest.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Bala, A., Kaushal, R. Jaywalking detection and localization in street scene videos using fine-tuned convolutional neural networks. Multimed Tools Appl 82, 34771–34791 (2023). https://doi.org/10.1007/s11042-023-14922-z
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-023-14922-z