Multi-task Learning Model for Detecting and Filtering Internet Violent Images for Children
Mô hình đa nhiệm nhận diện và lọc hình ảnh bạo lực cho trẻ em
Abstract
The Internet has emerged as an essential daily information access, but exposing children to inappropriate content can impair their early development. Existing content filtering methods exhibit limitations in accurately and efficiently detecting diverse inappropriate internet content. In this paper, we propose a multi-task learning model for detecting and filtering violent images to provide safer online experiences. The multi-task model is developed from the pre-trained lightweight base model such as MobileNetv2 to enable proper integration within web browser extensions. Pure training to detect violent images could raise false alarms in the classification results when the landscape or object images don’t contain any human, hence we develop two joint learning tasks such as detecting humans and detecting violent images simultaneously. Our experiments demonstrate that the proposed multi-task approach with binary rule achieves 98.5% accuracy, outperforming the single-task model for detecting violent images by a margin. Thereafter, the multi-task model is also integrated into the web extension to detect and filter out violent images to prevent children from harmful content.
References
Unicef-annual-report-2021, https://www.unicef.org/reports/unicefannual-report-2021, Last accessed 21 Feb, 2024.
Van Bruwaene, D., Huang, Q. & Inkpen, D. "A multiplatform dataset for detecting cyberbullying in social media", Lang Resources & Evaluation, 54, 851–874, 2020.
Van Bruwaene, David, Qianjia Huang, and Diana Inkpen, "A multi-platform dataset for detecting cyberbullying in social media", Language Resources and Evaluation, 54, 851-874, 2020.
Ramzan, Muhammad, Adnan Abid, Hikmat Ullah Khan, Shahid Mahmood Awan, Amina Ismail, Muzamil Ahmed, Mahwish Ilyas, and Ahsan Mahmood, "A review on stateof-the-art violence detection techniques", IEEE Access, 7: 107560-107575, 2019.
Mumtaz, N., Ejaz, N., Habib, S., Mohsin, S. M., Tiwari, P., Band ,"An overview of violence detection techniques: current challenges and future directions", Artificial intelligence review, 56(5), 4641-4666, 2023
Netnanny, https://www.netnanny.com/, Last accessed 21 Feb, 2024
Cybersitter, https://www.cybersitter.com/, Last accessed 21 Feb, 2024
Maxprotect, https://www.maxprotect.com/, Last accessed 21 Feb, 2024
M. Hammami, Y. Chahir and L. Chen, “WebGuard: Web based adult content detection and filtering system," In Proceedings IEEE/WIC International Conference on Web Intelligence (WI 2003), Halifax, NS, Canada, pp. 574-578,
Hu, Weiming, Haiqiang Zuo, Ou Wu, Yunfei Chen, Zhongfei Zhang, and David Suter. “Recognition of adult images, videos, and web page bags", ACM Transactions on Multimedia Computing, Communications, and Applications (TOMM) 7, no. 1: 1-24, 2011
Sharma, Preeti, Manoj Kumar, and Hitesh Sharma, "Comprehensive analyses of image forgery detection methods from traditional to deep learning approaches: an evaluation," Multimedia Tools and Applications 82, no. 12,: 18117-18150, 2023.
Soliman, Mohamed Mostafa, Mohamed Hussein Kamal, Mina Abd El-Massih Nashed, Youssef Mohamed Mostafa, Bassel Safwat Chawky, and Dina Khattab, "Violence recognition from videos using deep learning techniques," In 2019 Ninth International Conference on Intelligent Computing and Information Systems (ICICIS), pp. 80-85, 2019.
Human Action Recognition (HAR) Dataset. https://www.kaggle.com/datasets/meetnagadia/humanaction-recognition-har-dataset, Last accessed 21 Feb, 2024
Caruana, "Rich Multitask learning," Machine learning, 28 (1997): 41-75.
Sandler, Mark, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, and Liang-Chieh Chen, "Mobilenetv2: Inverted residuals and linear bottlenecks," In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 4510-4520, 2018.
Google Chrome extension API, https://developer.chrome.com/docs/extensions, Last accessed 21 Feb, 2024.