More than just an auxiliary loss: Anti-spoofing Backbone Training via Adversarial Pseudo-depth Generation

  title={More than just an auxiliary loss: Anti-spoofing Backbone Training via Adversarial Pseudo-depth Generation},
  author={Chang Keun Paik and Naeun Ko and Young Joon Yoo},
In this paper, a new method of training pipeline is discussed to achieve significant performance on the task of anti-spoofing with RGB image. We explore and highlight the impact of using pseudo-depth to pre-train a network that will be used as the backbone to the final classifier. While the usage of pseudo-depth for anti-spoofing task is not a new idea on its own, previous endeavours utilize pseudodepth simply as another medium to extract features for performing prediction, or as part of many… 

Figures and Tables from this paper


Learning Deep Models for Face Anti-Spoofing: Binary or Auxiliary Supervision
This paper argues the importance of auxiliary supervision to guide the learning toward discriminative and generalizable cues, and introduces a new face anti-spoofing database that covers a large range of illumination, subject, and pose variations.
Learn Convolutional Neural Network for Face Anti-Spoofing
Instead of designing feature by ourselves, the deep convolutional neural network is relied on to learn features of high discriminative ability in a supervised manner and combined with some data pre-processing, the face anti-spoofing performance improves drastically.
Image-to-Image Translation with Conditional Adversarial Networks
Conditional adversarial networks are investigated as a general-purpose solution to image-to-image translation problems and it is demonstrated that this approach is effective at synthesizing photos from label maps, reconstructing objects from edge maps, and colorizing images, among other tasks.
Unsupervised Domain Adaptation for Face Anti-Spoofing
This work introduces an unsupervised domain adaptation face anti-spoofing scheme to address the real-world scenario that learns the classifier for the target domain based on training samples in a different source domain, and introduces a new database for face spoofing detection.
Deep Spatial Gradient and Temporal Depth Learning for Face Anti-Spoofing
A new approach to detect presentation attacks from multiple frames based on two insights, able to capture discriminative details via Residual Spatial Gradient Block (RSGB) and encode spatio-temporal information from Spatio-Temporal Propagation Module (STPM) efficiently.
An original face anti-spoofing approach using partial convolutional neural network
This work extracts the deep partial features from the convolutional neural network (CNN) to distinguish the real and fake faces and uses the block principle component analysis (PCA) method to reduce the dimensionality of features that can avoid the over-fitting problem.
FeatherNets: Convolutional Neural Networks as Light as Feather for Face Anti-Spoofing
  • P. Zhang, F. Zou, +5 authors K. Li
  • Computer Science
    2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)
  • 2019
An extreme light network architecture(FeatherNet A/B) is proposed with a streaming module which fixes the weakness of Global Average Pooling and uses less parameters and a novel fusion procedure with "ensemble + cascade" structure is presented to satisfy the performance preferred use cases.
Conditional Image Synthesis with Auxiliary Classifier GANs
A variant of GANs employing label conditioning that results in 128 x 128 resolution image samples exhibiting global coherence is constructed and it is demonstrated that high resolution samples provide class information not present in low resolution samples.
Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification
This work proposes a Parametric Rectified Linear Unit (PReLU) that generalizes the traditional rectified unit and derives a robust initialization method that particularly considers the rectifier nonlinearities.
CelebA-Spoof: Large-Scale Face Anti-Spoofing Dataset with Rich Annotations
This work contributes a large-scale face anti-spoofing dataset, CelebA-Spoof, with the following appealing properties, and carefully benchmark existing methods in a unified multi-task framework, Auxiliary Information Embedding Network (AENet), and reveal several valuable observations.