Artificial Intelligence for Detecting Media Piracy

Bibliographic Reference(s)

• G. Zhu and D. Doermann, “Automatic Document Logo Detection,” Proc. Conf. Doc. Analysis Recogn., Vol. 2, pp. 864 –868, 2007. EXTERNAL
• O. M. Parkhi, A. Vedaldi, and A. Zisserman, “Deep Face Recognition,” Proc. Bri. Mach. Vision Conf., 2015. EXTERNAL
• S. M. M. Dezfooli et al., “Universal Adversarial Perturbations,” 2016. [Online]. Available: http://arxiv.org/abs/1610.08401 EXTERNAL
• K. Li et al., “Logo Detection with Extendibility and Discrimination,” Multimed Tools Appl., 72(2):1285–1310, 2014. EXTERNAL
• D. G. Lowe, “Object Recognition from Local Scale-Invariant Features,” Proc. Conf. Comput. Vision, Vol. 2, pp. 1150–1157, 1999. EXTERNAL
• A. Karpathy et al., “Large-Scale Video Classification with Convolutional Neural Networks,” Proc. IEEE Conf. Comput. Vision Patt. Recogn., pp. 1725–1732, 2014. EXTERNAL
• A. D. Bagdanov et al., “Trademark Matching and Retrieval in Sports Video Databases,” Proc. Workshop Multimedia Info. Retri., pp. 79–86, 2007. EXTERNAL
• A. S. Razavian, H. Azizpour, J. Sullivan, and S. Carlsson, “CNN Features Off-the-Shelf: An Astounding Baseline for EXTERNAL
• J. Yosinski et al., “How Transferable Are Features in Deep Neural Networks?,” Proc. Conf. Neur. Inform. Process. Syst., pp. 3320–3328, 2014. EXTERNAL
• S. C. H. Hoi et al., “Logo-Net: Large-Scale Deep Logo Detection and Brand Recognition with Deep Regi99on-Based Convolutional Networks,” 2015. [Online]. Available: http://arxiv.org/abs/1511.02462 EXTERNAL
• A. Krizhevsky, I. Sutskever, and G. E. Hinton, “Imagenet Classification with Deep Convolutional Neural Networks,” Adv. Neu. Info. Process. Syst., 25:1097–1105, 2012. EXTERNAL
• O. Russakovsky et al., “ImageNet Large Scale Visual Recognition Challenge,” Int. J. Comput.Vision, 115(3):211–252, 2015. EXTERNAL
• F. N. Iandola et al., “Deeplogo: Hitting Logo Recognition with the Deep Neural Network Hammer,” 2015. [Online]. Available: http://arxiv.org/abs/1510.02131 EXTERNAL
• R. Boia et al., “Logo Localization and Recognition in Natural Images using Homographic Class Graphs,” Mach. Vision Appl., 27(2):287–301, 2016. EXTERNAL
• D. Leporini, “Architectures and Protocols Powering Illegal Content Streaming over the Internet,” Proc. Int. Broadcasting Convention, p. 7, 2015. EXTERNAL
• K. He et al., “Deep Residual Learning for Image Recognition,” Proc. IEEE Conf. Comput. Vision Patt. Recogn., pp. 770–778, 2016. Conclusion The rapid rise of AI is disrupting many industries. Using AI for automating tasks previously done by human operators allows cost-efficient deployment and development of many large-scale applications that were unfeasible before. In this paper, we presented how AI can contribute to fighting piracy. We analyzed which aspects of AI are applicable in the implementation of large-scale content monitoring systems that track and identify illegally distributed content. We demonstrated how AI can be trained for automatically recognizing the original source of a video stream, by analyzing the logo of the content distributor from whom the content has been pirated. We see this as a building block into more complex content monitoring systems, which integrate multiple AI components for real-time monitoring of redistribution of copyrighted material. EXTERNAL