ABSTRACT
Counterfeiting of physical goods is a global problem amounting to nearly 7% of world trade. While there have been a variety of overt technologies like holograms and specialized barcodes and covert technologies like taggants and PUFs, these solutions have had a limited impact on the counterfeit market due to a variety of factors - clonability, cost or adoption barriers. In this paper, we introduce a new mechanism that uses machine learning algorithms on microscopic images of physical objects to distinguish between genuine and counterfeit versions of the same product. The underlying principle of our system stems from the idea that microscopic characteristics in a genuine product or a class of products (corresponding to the same larger product line), exhibit inherent similarities that can be used to distinguish these products from their corresponding counterfeit versions. A key building block for our system is a wide-angle microscopy device compatible with a mobile device that enables a user to easily capture the microscopic image of a large area of a physical object. Based on the captured microscopic images, we show that using machine learning algorithms (ConvNets and bag of words), one can generate a highly accurate classification engine for separating the genuine versions of a product from the counterfeit ones; this property also holds for "super-fake" counterfeits observed in the marketplace that are not easily discernible from the human eye. We describe the design of an end-to-end physical authentication system leveraging mobile devices, portable hardware and a cloud-based object verification ecosystem. We evaluate our system using a large dataset of 3 million images across various objects and materials such as fabrics, leather, pills, electronics, toys and shoes. The classification accuracy is more than 98% and we show how our system works with a cellphone to verify the authenticity of everyday objects.
- Alpvision. http://www.alpvision.com/cryptoglyph-covert-marking.html.Google Scholar
- Ambarella Video Security and Surveillance IP-Camera Solutions. http://www.ambarella.com/products/security-ip-cameras.Google Scholar
- Counterfeit goods are linked to terror groups - Business - International Herald Tribune. http://www.nytimes.com/2007/02/12/business/worldbusiness/12iht-fake.4569452.html.Google Scholar
- Counterfeiting Intelligence Bureau. http://www.iccwbo. org/products-and-services/fighting-commercial-crime/counterfeiting-intelligence-bureau/.Google Scholar
- Fabric Science kit. http://www.amazon.com/J-J-Pizzutos-Fabric-Science-Swatch/dp/1609013581.Google Scholar
- Ingenia technology ltd. http://www.ingeniatechnology.com/.Google Scholar
- IQStructures. http://www.iqstructures.com/.Google Scholar
- Neenah Paper. http://www.neenahpaper.com/.Google Scholar
- NovaVision. http://www.novavisioninc.com/.Google Scholar
- Prooftag. http://www.prooftag.net/.Google Scholar
- skimage: Image processing in Python. http://scikit-image.org /.Google Scholar
- Sproxil. http://sproxil.com/.Google Scholar
- Tanneries Haas - cuir de veau - Barenia, novocalf. http://www.tanneries-haas.com/.Google Scholar
- TI DM368. http://www.ti.com/product/tms320dm368.Google Scholar
- Veho Discovery Deluxe 004. http://www.veho-uk.com/main/shop_detail.aspx?article=40.Google Scholar
- WHO Anti-counterfeiting technologies for protection of medicines. http://www. who.int/impact/events/IMPACT-ACTechnologiesv3LIS.pdf.Google Scholar
- Berg, T., Liu, J., Lee, S. W., Alexander, M. L., Jacobs, D. W., and Belhumeur, P. N. Birdsnap: Large-scale fine-grained visual categorization of birds. In Proc. Conf. Computer Vision and Pattern Recognition (CVPR) (June 2014). Google ScholarDigital Library
- Buchanan, J. D. R., Cowburn, R. P., Jausovec, A.-V., Petit, D., Seem, P., Xiong, G., Atkinson, D., Fenton, K., Allwood, D. A., and Bryan, M. T. Forgery: "Fingerprinting" documents and packaging. Nature 436 (July 2005), 475. Google ScholarCross Ref
- Caputo, B., and Jie, L. A performance evaluation of exact and approximate match kernels for object recognition. Electronic Letters on Computer Vision and Image Analysis 8, 3 (2009), 15--26.Google Scholar
- Clarkson, W., Weyrich, T., Finkelstein, A., Heninger, N. Halderman, A., and Felten, E. Fingerprinting blank paper using commodity scanners. In IEEE Security and Privacy (2009). Google ScholarDigital Library
- Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., and Fei-Fei, L. ImageNet: A Large-Scale Hierarchical Image Database. In CVPR09(2009).Google Scholar
- E. Metois, P. Yarin, N. S., and Smith, J. R. Fiberfingerprint identification. In Third Workshop on Automatic Identification (2002).Google Scholar
- et. al, B. J. Surface treated security paper and method and device for producing surface treated security paper, US Patent Number 5,193,854, 1993.Google Scholar
- et al, E. B. G. Coatings and ink designs for negotiable instruments, us patent number 6,155,604, 2000.Google Scholar
- Greene, E. B. Negotiable instrument, us patent number 4,634,148, 1987.Google Scholar
- He, K., Zhang, X., Ren, S., and Sun, J. Deep residual learning for image recognition. arXiv preprint arXiv:1512.03385 (2015).Google Scholar
- Jarrett, K., Kavukcuoglu, K., Ranzato, M., and LeCun, Y. What is the best multi-stage architecture for object recognition? In Computer Vision, 2009 IEEE 12th International Conference on (2009), IEEE, pp. 2146--2153. Google ScholarCross Ref
- Jia, Y., Shelhamer, E., Donahue, J., Karayev, S., Long, J., Girshick, R., Guadarrama, S., and Darrell, T. Caffe: Convolutional architecture for fast feature embedding. arXiv preprint arXiv:1408.5093(2014).Google Scholar
- Kimura, and Yoshihiro. Woven security label, us patent number 6,068,895,2000.Google Scholar
- Krizhevsky, A., Sutskever, I., and Hinton, G. E. Imagene classification with deep convolutional neural networks. In NIPS(2012), vol. 1, p. 4.Google Scholar
- Lazebnik, S., Schmid, C., and Ponce, J. Beyond bags of features: Spatial pyramid matching for recognizing natural scene categories. In Computer Vision and Pattern Recognition, 2006 IEEE Computer Society Conference on (2006), vol. 2, IEEE, pp. 2169--2178.Google ScholarDigital Library
- LeCun, Y., Bottou, L., Bengio, Y., and Haffner, P. Gradient-based learning applied to document recognition. Proceedings of the IEEE 86, 11 (1998), 2278--2324. Google ScholarCross Ref
- Maji, S., Kannala, J., Rahtu, E., Blaschko, M., and Vedaldi, A. Fine-grained visual classification of aircraft. Tech. rep., 2013.Google Scholar
- Sharma, A., Subramanian, L., and Brewer, E. A. Paperspeckle: microscopic fingerprinting of paper. In ACM Conference on Computer and Communications Security (2011), pp. 99--110. Google ScholarDigital Library
- Simonyan, K., and Zisserman, A. Very deep convolutional networks for large-scale image recognition. CoRR abs/1409.1556 (2014).Google Scholar
- Smith, J. R., and Sutherland, A. V. Microstructure based indicia. In Proceedings of the Second Workshop on Automatic Identification Advanced Technologies (New York, NY, USA, 1999), ACM, pp. 79--83.Google Scholar
- Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov D., Erhan, D., Vanhoucke, V., and Rabinovich, A. Going deeper with convolutions. In CVPR 2015 (2015). Google ScholarCross Ref
- van Renesse, R. L. Optical Document Security, Second Edition. Artech House, Inc, Norwood, MA, 1998.Google Scholar
- Vedaldi, A., and Fulkerson, B. VLFeat: An open and portable library of computer vision algorithms. http://www.vlfeat.org/, 2008.Google Scholar
- Zeiler, M. D., and Fergus, R. Visualizing and understanding convolutional networks CoRR abs/1311.2901 (2013).Google Scholar
- Zeira, Eitan; Ellett, D. Verification methods employing thermally--imageable substrates, us patent number 6107244, August 2000.Google Scholar
- Zhu, B., Wu, J., and Kankanhalli, M. S. Print signatures for document authentication. In ACM CCS '03 (New York, NY, USA, 2003), pp. 145--15 Google ScholarDigital Library
Index Terms
- The Fake vs Real Goods Problem: Microscopy and Machine Learning to the Rescue
Recommendations
Information Goods vs. Industrial Goods: Cost Structure and Competition
We study markets for information goods and find that they differ significantly from markets for traditional industrial goods. Markets for information goods in which products are vertically differentiated lack the segmentation inherent in markets for ...
Upgrading Information Security and Protection for Palm-Print Templates
AbstractBiometric systems proven to be one of the most reliable and robust method for human identification. Integration of biometrics among the standard of living provokes the necessity to vogue secure authentication systems. The use of palm-prints for ...
Branding Conspicuous Goods: An Analysis of the Effects of Social Influence and Competition
Branding decisions are critical for the success of new products. Prior research on branding and brand extension has primarily focused on how branding influences consumers' perceptions of product quality. However, consumers of conspicuous goods care not ...
Comments