Smoking Behavior Recognition Method Based on Improved YOLOV8
Keywords:
Improving YOLOV8, Smoking behavior recognition, Deformable colluvium, Attention mechanism, Residual attention, Nonmaximum suppressionAbstract
Cigarettes, as critical indicators of smoking behavior, often appear as small targets in video surveillance. Due to camera resolution limits and shooting distance, cigarettes may only appear as bars, increasing false detection possibilities. Additionally, complex environments or lighting can cause partial or complete occlusion of cigarette targets, complicating detection. To accurately and quickly identify smoking behavior in such conditions, an improved YOLOV8-based method is proposed to enhance real-time recognition performance. The improved YOLOV8 backbone network uses deformable convolution layers to replace some conventional layers, extracting the smoking behavior feature map. Deformable pooling layers are added to process and extract critical features. In the enhanced neck network, a residual attention module combines channel and spatial attention mechanisms to refine the critical feature map. The detection head network employs scale, spatial, and task awareness attention modules to fuse important features from different scales, locations, and tasks, respectively. This fusion determines the smoking behavior recognition frame and score, optimizing the recognition process. An improved loss function further refines the recognition box, boosting accuracy. Experiments demonstrate that this method effectively extracts critical features of smoking behavior, achieving better feature extraction and accurate identification across varying lighting conditions, resulting in high accuracy.
Downloads
References
Kazutaka, O. , Kentaro, T. , Ayumi, K. , Shunsuke, S. & Toshiki, S. (2022). Common brain cortical abnormality in smoking behavior and bipolar disorder: discriminant analysis using cortical thickness and surface area. Cerebral Cortex, 32(20),4386-4396.
Tahseen, S. & Danti, A. (2022). Relevance of psychophysiological and emotional features for the analysis of human behavior survey. ECS transactions, 107(1),607-626.
Jed E., R. , Suzanne, F. , David, C. , Alfred, S. , Susan, C. & James M., D. (2023). Smoking reduction using electronic nicotine delivery systems in combination with nicotine skin patches. Psychopharmacology, 240(9),1901-1909.
Lin, S. C. , Gathua, N. , Thompson, C. , Sripipatana, A. & Makaroff, L. (2022). Disparities in smoking prevalence and associations with mental health and substance use disorders in underserved communities across the United States. Cancer,128(9),1826-1831.
Hoon, J. S. , Jin, J. H. , Junhyun, K. & Eun-Cheol, P. (2021). Association between smoking behavior and insulin resistance using triglyceride-glucose index among South Korean adults. The Journal of Clinical Endocrinology & Metabolism,106(11),e4531-e4541.
Tran, D. N. , Nguyen, T. N. , Phi-Khanh, P. C. & Tran, D. T. (2021). An IoT-based design using accelerometers in animal behavior recognition systems. IEEE Sensors Journal, 22(18), 17515-17528.
Mahalakshmi, V. , Sandhu, M. , Shabaz, M. , Keshta, I. , Prasad, K. D. V. & Kuzieva, N. , et al. (2024). Few-shot learning-based human behavior recognition model. Computers in human behavior, 151(Feb.),1.1-1.12.
Chen, X. & Dinavahi, V. (2021). Group behavior pattern recognition algorithm based on spatio-temporal graph convolutional networks. Scientific Programming,2021(Pt.4),2934943.1-2934943.8.
Farooq, M. U. , Saad, M. N. M. & Khan, S. D. (2022). Motion-shape-based deep learning approach for divergence behavior detection in the high-density crowd. The visual computer, 38(5),1553-1577.
De Ocampo, A. L. P., & Dadios, E. (2019). Radial greed algorithm with rectified chromaticity for anchorless region proposal applied in aerial surveillance. International Journal of Advances in Intelligent Informatics, 5(3), 193-205.
Pham, T. N. , Nguyen, V. H. & Huh, J. H. (2023). Integration of improved yolov5 for face mask detector and auto-labeling to generate dataset for fighting against COVID-19. The Journal of Supercomputing,79(8),8966-8992.
Manssor, S. A. F. , Sun, S. , Abdalmajed, M. & Ali, S. (2022). Real-time human detection in thermal infrared imaging at night using the enhanced tiny-yolov3 network. Journal of Real-Time Image Processing, 19(2), 261-274.
Tan, G. X., Cen, M. W. & Su, R. J. (2023). Design of Vehicle and Pedestrian Detection Network Based on YOLOv4.Computer Simulation,40(4):128-133.
Joris Guérin, Stéphane Thiery, Nyiri, E. , Gibaru, O. & Boots, B. (2021). Combining pre-trained cnn feature extractors to enhance the clustering of complex natural images. Neurocomputing, 423(Jan.29), 551-571.
A. L. P. De Ocampo, "Normalized Difference Vegetation Index (NDVI) Estimation based on Filter Augmented Imaging," 2023 International Electrical Engineering Congress (iEECON), Krabi, Thailand,2023,pp.84-88, doi: 10.1109 IEECON56657.2023.10126616.
Sreelakshmi, D. & Inthiyaz, S. (2021). Fast and denoise feature extraction based admf–cnn with a global framework for MRI brain image. International Journal of Speech Technology, 24(3), 1-16.
Sabry, E. S., Elagooz, S., El-Samie, F. E. A., El-Bahnasawy, N. A. & El-Banby, G. (2024). Assessment of various feature extraction methods for object discrimination in different scenarios. Journal of Optics, 53(1),49-69.
Schleider, L. , Pasiliao, E. L. , Qiang, Z. & Zheng, Q. P. (2022). A study of feature representation via neural network feature extraction and weighted distance for clustering. Journal of Combinatorial Optimization, 44(4),3083-3105.
Lahoti, G. , Ranjan, C. , Chen, J. , Yan, H. & Zhang, C. (2023). Convolutional neural network-assisted adaptive sampling for sparse feature detection in image and video data. IEEE Intelligent Systems,38(1),45-57.
Damaneh, M. M., Mohanna, F. & Jafari, P. (2022). Static hand gesture recognition in sign language based on a convolutional neural network with feature extraction method using orb descriptor and Gabor filter. Expert Syst. Appl, 211(Jan.), 118559.1-118559.13.
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
All papers should be submitted electronically. All submitted manuscripts must be original work that is not under submission at another journal or under consideration for publication in another form, such as a monograph or chapter of a book. Authors of submitted papers are obligated not to submit their paper for publication elsewhere until an editorial decision is rendered on their submission. Further, authors of accepted papers are prohibited from publishing the results in other publications that appear before the paper is published in the Journal unless they receive approval for doing so from the Editor-In-Chief.
IJISAE open access articles are licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. This license lets the audience to give appropriate credit, provide a link to the license, and indicate if changes were made and if they remix, transform, or build upon the material, they must distribute contributions under the same license as the original.
