Replayed Video Attack Detection Based on Motion Blur Analysis

Abstract

Face presentation attacks are main threats to face recognition system, and many presentation attack detection (PAD) methods have been proposed in recent few years. Although these methods have achieved significant performance in some specific intrusion modes, difficulties still exist in addressing replayed video attacks. Thats because replayed fake faces contain a variety of aliveness signals such as eye blinking and facial expression changes. Replayed video attacks occurred when attackers try to invade biometric systems by presenting face videos in front of cameras, and these videos are often launched by a liquidcrystal display (LCD) screen. Due to the smearing effects and movements of LCD, videos captured from real and replayed fake faces present different motion blurs, which mainly reflected in blur intensity variation and blur width. Based on these descriptions, a motion blur analysis based method is proposed to deal with replayed video attack problem. We first present a 1D convolutional neural network (CNN) for motion blur intensity variation description in time domain, which consists of a serial of 1D convolutional and pooling filters. Then, a local similar pattern (LSP) feature is introduced to extract blur width. Finally, features extracted from 1D CNN and LSP are fused to detect replayed video attacks. Extensive experiments on two standard face PAD databases, i.e., Relay-Attack and OULU-NPU, indicate that our proposed method based on motion blur analysis significantly outperforms the state-of-the-art methods and show excellent generalization capability.

Multi-Perspective Dynamic Features for Cross-Database Face Presentation Attack Detection

“…However, this algorithm was tested exclusively on a single database, and its cross-database performance is unknown. In [16], Li et al analyzed the effect on motion blur information caused by screen smearing in replay attack, and it is effective in detecting replay attack. The slight downside is that the motion of the face itself is not being exploited.…”

“…In video attacks, secondary imaging degrades the facial motion patterns, which differ significantly from those of a genuine face. Therefore, motion patterns can be used as a feature for face presentation attack detection [16], [17].…”

A Kernel Density Estimation Based Loss Function and its Application to ASV-Spoofing Detection

“…DNN-based approaches have been widely explored for audio- [13] and video-based anti-spoofing [14], [15]. Three key points are important for building a DNN-based antispoofing system with generalization capabilities: (i) architecture, (ii) input features, and (iii) loss function.…”

“…Multiple types of DNN architectures have been explored, such as feedforward DNN [16], convolutional neural network (CNN) [16], [17], recurrent neural network (RNN) [17], [18], gated recurrent neural network (GRCNN) [19], light convolutional gated recurrent neural network (LC-GRNN) [9], light convolutional neural network (LCNN) [20], central difference convolutional network (CDCN) [14], etc. Also, a wide range of features have been proposed to train these models, such as spectrogram [12], linear frequency cepstral coefficients (LFCC) [21], constant Q cepstral coefficients (CQCC) [22], raw speech samples [23], local similar pattern (LSP) features [15], signal-to-noise mask (SNM) [19] features, etc. Normally, the architecture of the DNN is adapted to the dimension of the input features, and viceversa.…”

Secure sliding mode control of interval type‐2 fuzzy systems against intermittent denial‐of‐service attacks

“…This indicates the significance of protecting cyberlayers against various attacks. In this regard, a lot of efforts have been made from the perspective of attack detection 4 and security control subject to cyberattacks 5,6 . For the latter, typical cyberattacks usually refer to: denial‐of‐service (DoS) attacks, deception attacks, and replay attacks 7,8 .…”