Exploring the Vulnerabilities of PRNU-based Camera Fingerprinting
Photos taken with a digital camera carry imperfections imperceptible to the human eye, but which can be deduced using image processing techniques. It is possible to link these imperfections to a single camera, which raises privacy concerns. Watch this video to learn about our approach to suppress this camera-identifying information from images.
S. Banerjee and A. Ross, “Smartphone Camera De-identification while Preserving Biometric Utility,” Proc. of 10th IEEE International Conference on Biometrics: Theory, Applications and Systems (BTAS), (Tampa, USA), September 2019.
S. Banerjee, V. Mirjalili, A. Ross, “Spoofing PRNU Patterns of Iris Sensors while Preserving Iris Recognition,” Proc. of 5th IEEE International Conference on Identity, Security and Behavior Analysis (ISBA), (Hyderabad, India), January 2019.
Iris Anti-Spoofing: Detecting Presentation Attacks
Iris biometric systems are vulnerable to “spoofing,” where adversaries attempt to trick the system with a fake iris. Watch this video to learn about our approach to defending against these attacks!
S. Hoffman, R. Sharma, A. Ross, “Iris + Ocular: Generalized Iris Presentation Attack Detection Using Multiple Convolutional Neural Networks,” Proc. of 12th IAPR International Conference on Biometrics (ICB), (Crete, Greece), June 2019.
S. Hoffman, R. Sharma, A. Ross, "Convolutional Neural Networks for Iris Presentation Attack Detection: Toward Cross-Dataset and Cross-Sensor Generalization," Proc. of IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), (Salt Lake City, USA), June 2018.
Spoofing Faces Using Makeup
Cosmetic makeup maybe applied to a person’s face with the intent to impersonate another person and fool an automated biometric recognition system.
C. Chen, A. Dantcheva, T. Swearingen, A. Ross, "Spoofing Faces Using Makeup: An Investigative Study," Proc. of 3rd IEEE International Conference on Identity, Security and Behavior Analysis (ISBA 2017), (New Delhi, India), February 2017.
Fusing Multiple AutoML Models Using Score Fusion: A Case Study in Medical Image Classification
Identifying stem cells in MRI scans is considered a difficult and time consuming task. In the video, we consider automated machine learning tools for identifying stem cells.
T. Swearingen, W. Drevo, B. Cyphers, A. Cuesta-Infante, A. Ross, K. Veeramachaneni, “ATM: A Distributed, Collaborative, Scalable System for Automated Machine Learning,” IEEE International Conference on Big Data, pp. 151 - 162, (Boston, USA), December 2017.
M. J. Afridi, A. Ross, E. Shapiro, "L-CNN: Exploiting Labeling Latency in a CNN Learning Framework," Proc. of 23rd International Conference on Pattern Recognition (ICPR), (Cancun, Mexico), December 2016.
DeepTalk: Vocal style transfer for realistic speech synthesis
Speech synthesis, the generation of artificial speech from an input, remains a difficult and well motivated task. In this video, we discuss how deep learning based natural language processing can be used to help generate realistic speech. For more, see: https://comartsci.msu.edu/nextgen-media-innovation-lab/nabpilot