A Review on Human Activity Recognition System


EOI: 10.11242/viva-tech.01.05.002

Download Full Text here



Citation

Priya Pathak, Vinod Choudhari, Mansi Patil, Bhavika Thakur, "A Review on Human Activity Recognition System", VIVA-IJRI Volume 1, Issue 6, Article 2, pp. 1-5, 2023. Published by Computer Engineering Department, VIVA Institute of Technology, Virar, India.

Abstract

Human activity recognition (HAR) is the process of interpreting human motion using computer and machine vision technology. Gestures, behaviours, and activities which are recorded by sensors are used to detect human activity. HAR is an active research area combining Convolutional Neural Networks (CNN) and feature extraction classification methods for surveillance and other applications. However, accurately identifying HAR from a sequence of frames is challenging due to cluttered backgrounds, different viewpoints, low resolution, and partial occlusion. Current CNN-based techniques require large computational classifiers along with convolutional networks having local receptive fields which limits the performance to capture long-range temporal information. Therefore, this work introduces a low computational power approach for accurate HAR, which overcomes the problems mentioned above and accurately encodes relative spatial information. In this proposed method, the You Only Look Once (YOLO) network is utilized as a backbone CNN model. For training the model, we constructed a large dataset of videos by labelling each frame with a set of activities and positions.

Keywords

YOLO, HAR, CNN, Deep Learning, Classification

References

  1. I. M. Nasir, M. Raza, J. H. Shah, M. Attique Khan and A. Rehman, "Human Action Recognition using Machine Learning in Uncontrolled Environment" IEEE, 2021 1st International Conference on Artificial Intelligence and Data Analytics (CAIDA).
  2. S. K et al., "Machine Learning-based Human Activity Recognition using Neighbourhood Component Analysis" IEEE, 2021 5th International Conference on Computing Methodologies and Communication (ICCMC).
  3. M. M. Hossain Shuvo, N. Ahmed, K. Nouduri and K. Palaniappan, "A Hybrid Approach for Human Activity Recognition with Support Vector Machine and 1D Convolutional Neural Network," 2020 IEEE Applied Imagery Pattern Recognition Workshop (AIPR).
  4. J. D. Domingo, J. Gómez-García-Bermejo and E. Zalama, "Improving Human Activity Recognition Integrating LSTM With Different Data Sources: Features, Object Detection and Skeleton Tracking," in IEEE Access, vol. 10.
  5. R. Saini and V. Maan, "Human Activity and Gesture Recognition: A Review," 2020 International Conference on Emerging Trends in Communication, Control and Computing (ICONC3), IEEE 2020.
  6. L. Xie, J. Tian, G. Ding and Q. Zhao, "Human activity recognition method based on inertial sensor and barometer," 2018 IEEE International Symposium on Inertial Sensors and Systems (INERTIAL).
  7. Md. Atikuzzaman, "Human activity recognition system from different poses with CNN”, 2020 Researchgate.
  8. Kai-Tai Song and Wei-Jyun Chen, "Human activity recognition using a mobile camera," 2011 8th International Conference on Ubiquitous Robots and Ambient Intelligence (URAI), 2011.
  9. L. Alpoim, A. F. da Silva and C. P. Santos, "Human Activity Recognition Systems: State of Art,"2019 IEEE 6th Portuguese Meeting on Bioengineering (ENBENG), 2019.
  10. D. Ravi, C. Wong, B. Lo and G. -Z. Yang, "Deep learning for human activity recognition: A resource-efficient implementation on low-power devices," 2016 IEEE 13th International Conference on Wearable and Implantable Body Sensor Networks (BSN), 2016.
  11. https://www.crcv.ucf.edu/research/data-sets/ucf50/, last accessed on 07/10/2022.