Face Spoof Detection With Image Distortion Analysis

By Di Wen, Hu Han, Anil K. Jain. IEEE Transactions on Information Forensics and Security, 2015

913 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Signal Processing

View Paper: https://www.semanticscholar.org/paper/cdac436dcebe8b2c90a8de5479bd3bbd8d9a087f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Xiaogang Wang
2. Luc Van Gool
3. Xiaoming Liu
4. Chen Change Loy
5. Dacheng Tao
6. Yu Qiao
7. Yi Yang
8. Xilin Chen

In Signal Processing

1. Xiaoming Liu
2. Yaojie Liu
3. Amin Jourabloo
4. Sébastien Marcel
5. Xilin Chen
6. Shiguang Shan
7. Hu Han
8. Zitong Yu

Citations in the corpus

• Secure Face Unlock: Spoof Detection on Smartphones (2016, methodology): “While face spoof detection under the smartphone unlock scenarios was studied in [20], it used a database containing only 50 subjects (videos of only 35 subjects are publicly available).”
• CelebA-Spoof: Large-Scale Face Anti-Spoofing Dataset with Rich Annotations (2020, background): “Second is the single-modal dataset, such as Replay Attack [5], CASIA-MFSD [39], MSU-MFSD [34], MSU-USSA [29] and HKBU-MARS V2 [21].”
• DeepFakes: a New Threat to Face Recognition? Assessment and Detection (2018, methodology): “The systems based on image quality measures (IQM) are borrowed from the domain of presentation (including replay attacks) attack detection, where such systems have shown good performance [12], [13].”
• Face Spoofing Detection Using Colour Texture Analysis (2016, methodology): “lack of high frequency information [2], [3], [4], [5], [6].”
• Single-Side Domain Generalization for Face Anti-Spoofing (2020, methodology): “This is because all other face anti-spoofing methods [6, 9, 23, 26, 39, 42] except for the MADDG [33] method pay no attention to the intrinsic distribution relationship among different domains.”
• Deep learning for deepfakes creation and detection: A survey (2019, background): “stems based on VGG [38] and Facenet [31,39] are unable to detect deepfakes eectively. Other methods such as lip-syncing approaches [40{42] and image quality metrics with support vector machine (SVM) [43,44] produce very high error rate when applied to detect deepfake videos from this newly produced ”
• Cross-Domain Face Presentation Attack Detection via Multi-Domain Disentangled Representation Learning (2020, methodology): “…that leads to poor generalization ability of PAD models under new application scenarios. face PAs, which assume that there are inherent disparities between live and spoof faces, such as texture in color space [4, 6], image distortion [48], temporal variation [39] or deep semantic features [50, 36].”
• Deep Learning for Deepfakes Creation and Detection (2019, methodology): “Other methods such as lip-syncing approaches [40–42] and image quality metrics with support vector machine (SVM) [43, 44] produce very high error rate when applied to detect deepfake videos from this newly produced data set.”
• Face Anti-Spoofing: Model Matters, so Does Data (2019, background): “Face anti-spoofing [40, 1, 15, 14] is an important, yet challenging problem in the face recognition community.”
• Face Anti-Spoofing with Human Material Perception (2020, methodology): “Six databases including OULU-NPU [58], SiW [14], CASIA-MFSD [59], Replay-Attack [60], MSU-MFSD [61] and SiW-M [29] are used in our experiments.”
• CASIA-SURF: A Large-Scale Multi-Modal Benchmark for Face Anti-Spoofing (2019, background): “One of the key points of this success is the availability of face anti-spoofing benchmark datasets [8], [9], [11], [12], [14], [15].”
• Efficient Acceleration of Deep Learning Inference on Resource-Constrained Edge Devices: A Review (2023, background): “For instance, face recognition is being actively used in door unlocking, mobile banking [383], device unlocking [384], and so on.”
• OULU-NPU: A Mobile Face Presentation Attack Database with Real-World Variations (2017, background): “Although the CASIA-FASD, is smaller than the Replay-Attack Database, it has shown to be more challenging benchmark dataset due to the diversity in the data, including attack types and (lesscontrolled) acquisition conditions in general, e.g. standoff distance and input sensor quality.”
• Multi-Adversarial Discriminative Deep Domain Generalization for Face Presentation Attack Detection (2019, methodology): “We evaluate our work on four public face anti-spoofing datasets, which contain both print and video replay attacks: Oulu-NPU [4] (O for short), CASIA-MFSD [35] (C for short), Idiap Replay-Attack [8] (I for short), and MSUMFSD [31] (M for short).”
• Deep Spatial Gradient and Temporal Depth Learning for Face Anti-Spoofing (2020, background): “Therefore, not only the research community but also the industry has recognized face anti-spoofing [18, 19, 4, 33, 39, 11, 23, 55, 1, 29, 12, 49, 45, 54, 21] as a critical role in securing the face recognition system.”
• Searching Central Difference Convolutional Networks for Face Anti-Spoofing (2020, methodology): “Method CASIA-MFSD [66] Replay-Attack [13] MSU-MFSD [57] Overall Video Cut Photo Wrapped Photo Video Digital Photo Printed Photo Printed Photo HR Video Mobile Video OC-SVMRBF +BSIF [3] 70.”
• Learning Deep Models for Face Anti-Spoofing: Binary or Auxiliary Supervision (2018, methodology): “Existing face anti-spoofing databases, such as NUAA [43], CASIA [50], Replay-Attack [17], and MSU-MFSD [45], were collected 3 − 5 years ago.”
• Deep Anomaly Detection for Generalized Face Anti-Spoofing (2019, methodology): “From the experimental results the paper concludes that: i) anomaly detection based systems are comparable to two classes based systems; and ii) neither of the two approaches perform well enough in the evaluated datasets (CASIA-FASD [39], Replay-Attack [8] and MSU-MFSD [33]).”