FeatherNets: Convolutional Neural Networks as Light as Feather for Face Anti-Spoofing

Zhang, Peng; Zou, Fang; Wu, Zhiwen; Dai, Nengli; Mark, Skarpness; Fu, Michael J.; Zhao, Juan; Li, Kai

doi:10.1109/cvprw.2019.00199

Cited by 64 publications

(45 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [48], patches with fixed-size are randomly extracted from cropped face areas, which can strength the ability of CNNs for perceiving spoof clues on the texture attribute other than the whole facial structure attribute. However, it is timeconsuming to traverse all patches for network inference.…”

Section: Multiscale Patch and Weighted Fine-tuningmentioning

confidence: 99%

Data-Fusion-Based Two-Stage Cascade Framework for Multimodality Face Anti-Spoofing

Liu

Wei²,

Lei

et al. 2022

IEEE Trans. Cogn. Dev. Syst.

View full text Add to dashboard Cite

Existing face anti-spoofing models using deep learning for multi-modality data suffer from low generalization in the case of using variety of presentation attacks such as 2D printing and high-precision 3D face masks. One of the main reasons is that the non-linearity of multi-spectral information used to preserve the intrinsic attributes between a real and a fake face are not well extracted. To address this issue, we propose a multi-modility data based two-stage cascade framework for face anti-spoofing. The proposed framework has two advantages. Firstly, we design a two-stage cascade architecture that can selectively fuse lowlevel and high-level features from different modalities to improve feature representation. Secondly, we use multi-modality data to construct a distance-free spectral on RGB and infrared (IR) to augment the non-linearity of data. The presented data fusion strategy is different from popular fusion approaches, since it can strengthen discrimination ability of network models on physical attribute features than identity structure features under certain constraints. In addition, a multi-scale patch based weighted finetuning strategy is designed to learn each specific local face region. Experimental results show that the proposed framework achieves better performance than other state-of-the-art methods on both benchmark datasets and self-established datasets, especially on multi-material masks spoofing.

show abstract

Section: Multiscale Patch and Weighted Fine-tuningmentioning

confidence: 99%

Data-Fusion-Based Two-Stage Cascade Framework for Multimodality Face Anti-Spoofing

Liu

Wei²,

Lei

et al. 2022

IEEE Trans. Cogn. Dev. Syst.

View full text Add to dashboard Cite

show abstract

“…Learning: Because of its popularity, deep learning techniques are also considered effective in solving the issue of building an anti-spoofing system. In the deep learning technique, a convolution neural network is used to detect spoofing attacks [9]. The process takes an input image, a process that consists of neurons that have a function of weight, bias, and activation, then classify it into certain categories.…”

Section: ) Deepmentioning

confidence: 99%

A Review on Face Anti-Spoofing

Perdana

Ardiyanto

Nugroho

2021

IJITEE

View full text Add to dashboard Cite

The biometric system is a security technology that uses information based on a living person's characteristics to verify or recognize the identity, such as facial recognition. Face recognition has numerous applications in the real world, such as access control and surveillance. But face recognition has a security issue of spoofing. A face anti-spoofing, a task to prevent fake authorization by breaching the face recognition systems using a photo, video, mask, or a different substitute for an authorized person's face, is used to overcome this challenge. There is also increasing research of new datasets by providing new types of attack or diversity to reach a better generalization. This paper review of the recent development includes a general understanding of face spoofing, anti-spoofing methods, and the latest development to solve the problem against various spoof types.

show abstract

“…ResNets [32], DenseNets [33] and most light networks, like MobileNetV2 [34], ShuffleNetV2 [35]. For face tasks, however, researchers [36], [37] have observed that CNNs with GAP layer are less accurate than those without GAP. Further, FeatherNets [37] replace the GAP with specially designed Streaming Module, and achieve significant performance for face anti-spoofing.…”

Section: ) the Weakness Of Gap In Fldmentioning

confidence: 99%

“…For face tasks, however, researchers [36], [37] have observed that CNNs with GAP layer are less accurate than those without GAP. Further, FeatherNets [37] replace the GAP with specially designed Streaming Module, and achieve significant performance for face anti-spoofing. Inspired by these works, we carry out experiments to evaluate the weakness of GAP in FLD and get the same conclusion ( Table 8).…”

Section: ) the Weakness Of Gap In Fldmentioning

confidence: 99%

FLDNet: Light Dense CNN for Fingerprint Liveness Detection

et al. 2020

View full text Add to dashboard Cite

Fingerprint liveness detection has gained increased attention recently due to the growing threat of spoof presentation attacks. Among the numerous attempts to deal with this problem, the Convolutional Neural Networks (CNN) based methods have shown impressive performance and great potential. However, there is a need for improving the generalization ability and reducing the complexity. Therefore, we propose a lightweight (0.48M parameters) and efficient network architecture, named FLDNet, with an attention pooling layer which overcomes the weakness of Global Average Pooling (GAP) in fingerprint anti-spoofing tasks. FLDNet consists of modified dense blocks which incorporate the residual path. The designed block architecture is compact and effectively boosts the detection accuracy. Experimental results on two datasets, LivDet 2013 and 2015, show the proposed approach achieves state-of-the-art performance in intra-sensor, cross-material and cross-sensor testing scenarios. For example, on LivDet 2015 dataset, FLDNet achieves 1.76% Average Classification Error (ACE) over all sensors and 3.31% against unkown spoof materials compared to 2.82% and 5.45% achieved by state-of-the-art methods. INDEX TERMS Fingerprint liveness detection, convolutional neural networks, presentation attacks, attention pooling.

show abstract

FeatherNets: Convolutional Neural Networks as Light as Feather for Face Anti-Spoofing

Cited by 64 publications

References 32 publications

Data-Fusion-Based Two-Stage Cascade Framework for Multimodality Face Anti-Spoofing

Data-Fusion-Based Two-Stage Cascade Framework for Multimodality Face Anti-Spoofing

A Review on Face Anti-Spoofing

FLDNet: Light Dense CNN for Fingerprint Liveness Detection

Contact Info

Product

Resources

About