DenseNet121 and Transfer Learning Based Video Analysis Method for Fight Detection in Streaming Images

Öz

Early detection of sudden fight incidents and prompt, effective intervention are crucial for ensuring public safety and preventing potential adverse situations. With the advancements in modern technology, video-based automatic fight detection systems enable security forces and authorities to take proactive measures by providing timely alerts. In this study, a deep learning-based method is proposed to accurately classify fight and non-fight situations. Specifically, video-based fight detection is performed using transfer learning with the DenseNet121 model, which has demonstrated successful results in image processing tasks. In the experimental studies, the widely used Peliculas and Hockey datasets from the literature were evaluated. The proposed method achieved 100% accuracy on the Peliculas dataset and 99.32% accuracy on the Hockey dataset. These high success rates indicate that the proposed method can be effectively used for fight detection.

Anahtar Kelimeler

• Peliculas dataset
• Hockey dataset
• Keyframe extraction

Akan Görüntülerde Kavga tespiti için DenseNet121 ve Transfer Öğrenme Tabanlı Video Analiz Yöntemi

Öz

Ani gelişen kavga olaylarının erken tespiti ve bu olaylara hızlı ve etkili bir şekilde müdahale edilmesi, kamu güvenliğinin sağlanması ve olası olumsuz durumların önlenmesi açısından büyük bir öneme sahiptir. Günümüz teknolojik imkanları doğrultusunda, video tabanlı otomatik kavga tespit sistemleri, güvenlik güçlerine ve yetkililere zamanında uyarılar sağlayarak proaktif önlemler alınmasını mümkün kılmaktadır. Bu çalışmada, kavga (fight) ve kavga içermeyen (noFight) durumların doğru bir şekilde sınıflandırılabilmesi amacıyla derin öğrenme tabanlı bir yöntem önerilmektedir. Özellikle, görüntü işleme alanında başarılı sonuçlar elde eden DenseNet121 modeli kullanılarak transfer öğrenme yöntemiyle video tabanlı kavga tespiti gerçekleştirilmiştir. Deneysel çalışmalarda, literatürde yaygın olarak kullanılan Peliculas ve Hockey veri setleri değerlendirilmiş olup, önerilen yöntem ile Peliculas veri seti üzerinde %100 doğruluk (Accuracy), Hockey veri seti üzerinde ise %99.32 doğruluk elde edilmiştir. Elde edilen yüksek başarı oranları, önerilen yöntemin kavga tespitinde etkin bir şekilde kullanılabileceğini göstermektedir.

Anahtar Kelimeler

• Peliculas veri seti
• Hockey veri seti
• Anahtar kare çıkarımı

Kaynakça

1. Akti, S., Ofli, F., Imran, M., & Ekenel, H. K. (2022, January). Fight detection from still images in the wild. In 2022 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW) (pp. 550–559). IEEE. https://doi.org/10.1109/WACVW54805.2022.00061
2. Akti, S., Tataroglu, G. A., & Ekenel, H. K. (2019, November). Vision-based fight detection from surveillance cameras. In 2019 Ninth International Conference on Image Processing Theory, Tools and Applications (IPTA) (pp. 1–6). IEEE. https://doi.org/10.1109/IPTA.2019.8936070
3. Akash, S. A., Moorthy, R. S. S., Esha, K., & Nathiya, N. (2022). Human violence detection using deep learning techniques. Journal of Physics: Conference Series, 2318(1), 012003. https://doi.org/10.1088/1742-6596/2318/1/012003
4. Ali, Y. Y., & Yaman, O. (2024). Analysis of LSTM, BiLSTM, and CNN methods for environmental sound identification in smart cities. In Proceedings of the 2024 8th International Artificial Intelligence and Data Processing Symposium (IDAP) (pp. 1–6). IEEE. https://doi.org/10.1109/IDAP64064.2024.10711138
5. Albawi, S., Mohammed, T. A. S., & Al-Zawi, S. (2018). Understanding of a convolutional neural network. In 2017 International Conference on Engineering and Technology (ICET) (pp. 1–6). IEEE. https://doi.org/10.1109/ICEngTechnol.2017.8308186
6. Bayram, B., Kunduracioglu, I., Ince, S., & Pacal, I. (2025). A systematic review of deep learning in MRI-based cerebral vascular occlusion-based brain diseases. Neuroscience, 568, 1–22. https://doi.org/10.1016/j.neuroscience.2025.01.020
7. Carneiro, S. A., Da Silva, G. P., Guimaraes, S. J. F., & Pedrini, H. (2019, October). Fight detection in video sequences based on multi-stream convolutional neural networks. In 2019 32nd SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI) (pp. 8–15). IEEE. https://doi.org/10.1109/SIBGRAPI.2019.00010
8. Demir, K., & Yaman, O. (2024). Projector deep feature extraction-based garbage image classification model using underwater images. Multimedia Tools and Applications, 83, 79437–79451. https://doi.org/10.1007/s11042-024-18731-w

9. Ertam, F., Kilincer, L. F., & Yaman, O. (2017). Intrusion detection in computer networks via machine learning algorithms. In Proceedings of the 2017 International Artificial Intelligence and Data Processing Symposium (IDAP) (pp. 1–4). IEEE. https://doi.org/10.1109/IDAP.2017.8090165
10. Gu, J., Wang, Z., Kuen, J., Ma, L., Shahroudy, A., Shuai, B., ... & Chen, T. (2018). Recent advances in convolutional neural networks. Pattern Recognition, 77, 354–377. https://doi.org/10.1016/j.patcog.2017.10.013
11. Hassner, T., Itcher, Y., & Kliper-Gross, O. (2012). Violent flows: Real-time detection of violent crowd behavior. In Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) (pp. 1–6). https://doi.org/10.1109/CVPRW.2012.6239348
12. Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017). Densely connected convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 4700–4708). https://doi.org/10.1109/CVPR.2017.243
13. Ince, S., Kunduracioglu, I., Algarnic, A., Bayram, B., & Pacal, I. (2025). Deep learning for cerebral vascular occlusion segmentation: A novel ConvNeXtV2 and GRN-integrated U-Net framework for diffusion-weighted imaging. Neuroscience, 541, 113–129. https://doi.org/10.1016/j.neuroscience.2025.04.010
14. Iqrar, W., Shahzad, A., Hameed, W., & Abidien, M. Z. (2023, January). A real-time sequence based human activity detection system. In 2023 International Multi-disciplinary Conference in Emerging Research Trends (IMCERT) (pp. 1–5). IEEE. https://doi.org/10.1109/IMCERT57083.2023.10075257
15. İnce, S., Kunduracioglu, I., Bayram, B., Pacal, I. (2025). U-Net-Based Models for Precise Brain Stroke Segmentation. Chaos Theory and Applications, 7(1), 50-60. https://doi.org/10.51537/chaos.1605529
16. Jaiswal, S. G., Mohod, S. W., & Sharma, D. (2023, September). HOG ensembled boosting machine learning approach for violent video classification. Indian Journal of Science and Technology, 16(34), 2709–2718. https://doi.org/10.17485/IJST/v16i34.1777
17. Karaduman, G., & Akin, E. (2020). A deep learning based method for detecting of wear on the current collector strips’ surfaces of the pantograph in railways. IEEE Access, 8, 183799–183812. https://doi.org/10.1109/ACCESS.2020.3029555
18. Kilincer, I. F., Ertam, F., & Sengur, A. (2021). Machine learning methods for cyber security intrusion detection: Datasets and comparative study. Computer Networks, 188, Article 107840. https://doi.org/10.1016/j.comnet.2021.107840
19. Kunduracioglu, I., & Pacal, I. (2024). Advancements in deep learning for accurate classification of grape leaves and diagnosis of grape diseases. Journal of Plant Diseases and Protection, 131, 1061–1080. https://doi.org/10.1007/s41348-024-00896-z
20. LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436–444. https://doi.org/10.1038/nature14539
21. Nadeem, M. S., & Kurugollu, F. (2021, October 25). Effectiveness of synthetic images in violence detection. TechRxiv. https://doi.org/10.36227/techrxiv.16822123.v1
22. Nievas, E. B., Suarez, O. D., Garcia, G. B., & Sukthankar, R. (2011). Violence detection in video using computer vision techniques. In Computer Analysis of Images and Patterns (CAIP 2011), Lecture Notes in Computer Science, vol 6854 (pp. 332–339). Springer. https://doi.org/10.1007/978-3-642-23678-5_39
23. Ozel, M. E., Apaydin, N. N., Yaman, O., & Karakose, M. (2025). A CNN-based method using optimized parameters for dynamic human action recognition. In Proceedings of the 2025 29th International Conference on Information Technology (IT) (pp. 1–4). IEEE. https://doi.org/10.1109/IT64745.2025.10930265
24. Pacal, I., Kunduracioglu, I., Alma, M. H., Deveci, M., Kadry, S., Nedoma, J., Slany, V., & Martinek, R. (2024). A systematic review of deep learning techniques for plant diseases. Artificial Intelligence Review, 57, 304. https://doi.org/10.1007/s10462-024-10944-7
25. Pacal, I., Ozdemir, B., Zeynalov, J., Gasimov, H., & Pacal, N. (2025). A novel CNN-ViT-based deep learning model for early skin cancer diagnosis. Biomedical Signal Processing and Control, 104, 107627. https://doi.org/10.1016/j.bspc.2025.107627
26. Pacal, I., & Attallah, O. (2025). Hybrid deep learning model for automated colorectal cancer detection using local and global feature extraction. Knowledge-Based Systems, 319, 113625. https://doi.org/10.1016/j.knosys.2025.113625
27. Pleiss, G., Chen, D., Huang, G., Li, T., van der Maaten, L., & Weinberger, K. Q. (2017). Memory‐efficient implementation of DenseNets (arXiv:1707.06990) [Preprint]. arXiv. https://doi.org/10.48550/arXiv.1707.06990
28. Stanley, B. F., R. K. R., J., & P. P., B. (2023, December). Efficient violence recognition system using spatio-temporal shift multi-scale attention model. In 2023 IEEE 20th India Council International Conference (INDICON) (pp. 509–514). IEEE. https://doi.org/10.1109/INDICON59947.2023.10440925
29. Venkatesh, S., Anand, A., G. S., Ramakrishnan, A., & Vijayaraghavan, V. (2020). Real-time surveillance based crime detection for edge devices. In Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (pp. 801–809). SCITEPRESS. https://doi.org/10.5220/0008990108010809
30. Yaman, O., Yetiş, H., & Karaköse, M. (2021). Image processing and machine learning-based classification method for hyperspectral images. The Journal of Engineering, 2021, 85–96. https://doi.org/10.1049/tje2.12012