Human Action Recognition Using Dense Trajectories

Dileep Labana; Kirit Modi

doi:10.17485/IJST/v16i43.2408

Article

Human Action Recognition Using Dense Trajectories

VIEWS 274
PDF 59

Indian Journal of Science and Technology

DOI: 10.17485/IJST/v16i43.2408

Year: 2023, Volume: 16, Issue: 43, Pages: 3846-3853

Original Article

Human Action Recognition Using Dense Trajectories

Dileep Labana^1*, Kirit Modi²

¹Ph.D. Student, Computer Science & Engineering, Parul Institute of Technology, Parul University, at & post Vaghodiya, Vadodara, 391760, Gujarat, India
²Professor, Computer Engineering Department, Sankalchand Patel University, Visnagar, 384315, Gujarat, India

*Corresponding Author
Email: [email protected]

Received Date:21 September 2023, Accepted Date:03 October 2023, Published Date:13 November 2023

This work is licensed under a Creative Commons Attribution 4.0 International License.

Abstract

Objective: To develop a robust and effective computer vision system that can automatically identify and classify human actions in video data, considering the temporal dynamics and various environmental conditions. This technology has numerous applications in surveillance, human-computer interaction, and video analysis. Methods: The key methods for dense trajectory extraction include the dense optical flow, which computes motion vectors for each point, and the use of key point detectors like the Scale-Invariant Feature Transform (SIFT) or the Harris corner detector. Findings: By describing the motion of the trajectories, trajectory descriptors produce remarkably strong results on their own, such as 90.2% on KTH and 47.7% on Hollywood2 for dense trajectories. This demonstrates the significance of the motion data present in the local trajectory patterns. Because the trajectory descriptors catch a lot of camera motion, we only report 67.2% on YouTube. Novelty: In this study, a method for modelling movies that combines dense sampling and feature tracking is presented. Compared to earlier video descriptions, our dense trajectories are more reliable. They effectively capture the motion data in the movies and outperform cutting-edge action categorization techniques in terms of performance.

Keywords: Human action recognition, Scale-Invariant feature transform, Histograms of oriented gradients, Spatial and temporal interest points, Histograms of optical flow

References

Agarwal S, Gupta MK. Context Aware Image Sentiment Classification using Deep Learning Techniques. Indian Journal Of Science And Technology. 2022;15(47):2619–2627. Available from: https://doi.org/10.17485/IJST/v15i47.1907
Patel CI, Labana D, Pandya S, Modi K, Ghayvat H, Awais M. Histogram of Oriented Gradient-Based Fusion of Features for Human Action Recognition in Action Video Sequences. Sensors. 2020;20(24):1–32. Available from: https://doi.org/10.3390/s20247299
Srilakshmi N, Radha N. An Enhancement of Deep Positional Attention-Based Human Action Recognition by Using Geometric Positional Features. Indian Journal Of Science And Technology. 2023;16(29):2190–2197. Available from: https://doi.org/10.17485/IJST/v16i29.379
Xu Y, Zhou F, Wang L, Peng W, Zhang K. Optimization of Action Recognition Model Based on Multi-Task Learning and Boundary Gradient. Electronics. 2021;10(19):1–16. Available from: https://doi.org/10.3390/electronics10192380
Patel R, Vaghela R, Chopade M, Patel P, Bhatt D. Integrated Neuroinformatics: Analytics and Application. In: Knowledge Modelling and Big Data Analytics in Healthcare (1). Boca Raton. CRC Press. 2021.
Labana D, Modi K. Human Action Recognition via Multi-Task Learning. Journal of Emerging Technologies and Innovative Research. 2023;10(7):409–414. Available from: https://www.jetir.org/papers/JETIR2307548.pdf
Papadopoulos K, Demisse G, Ghorbel E, Antunes M, Aouada D, Ottersten B. Localized Trajectories for 2D and 3D Action Recognition. Sensors. 2019;19(16):1–22. Available from: https://doi.org/10.3390/s19163503
Nguyen TT, Nguyen TP, Bouchara F. Directional dense-trajectory-based patterns for dynamic texture recognition. IET Computer Vision. 2020;p. 162–176. Available from: https://doi.org/10.1049/iet-cvi.2019.0455
Arif S, Ul-Hassan T, Hussain F, Wang J, Fei Z. Video representation by dense trajectories motion map applied to human activity recognition. International Journal of Computers and Applications. 2020;42(5):474–484. Available from: https://doi.org/10.1080/1206212X.2018.1486001
Morceli BDM, Poz APD. Road extraction from low-cost GNSS-device dense trajectories. Journal of Location Based Services. 2023;17(3):251–270. Available from: https://doi.org/10.1080/17489725.2023.2216670
Camarena F, Chang L, Gonzalez-Mendoza M, Cuevas-Ascencio RJ. Action recognition by key trajectories. Pattern Analysis and Applications. 2022;25(2):409–423. Available from: https://link.springer.com/content/pdf/10.1007/s10044-021-01054-z.pdf?pdf=button
Yi Y, Li A, Zhou X. Human action recognition based on action relevance weighted encoding. Signal Processing: Image Communication. 2020;80:115640. Available from: https://doi.org/10.1016/j.image.2019.115640
Zhao H, Dang J, Wang S, Wang Y, Gao D. Dense Trajectory Action Recognition Algorithm Based on Improved SURF. IOP Conference Series: Earth and Environmental Science. 2019;252(3):1–8. Available from: https://iopscience.iop.org/article/10.1088/1755-1315/252/3/032179/meta
Roseline V, Chellam GH. A Novel Fusion Attention Algorithm for Sentimental Image Analysis. Indian Journal of Science and Technology. 2022;15(9):386–394. Available from: https://doi.org/10.17485/IJST/v15i9.2159
Mefteh S, Kaâniche MBB, Ksantini R, Bouhoula A. A novel multispectral corner detector and a new local descriptor: an application to human posture recognition. Multimedia Tools and Applications. 2023;82:28937–28956. Available from: https://doi.org/10.1007/s11042-023-14788-1

Copyright

© 2023 Labana & Modi. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Published By Indian Society for Education and Environment (iSee)