Adaptive Transformers for Robust Few-shot Cross-domain Face Anti-spoofing

Huang, Hsin-Ping; Sun, Deqing; Liu, Yaojie; Chu, Wen-Sheng; Xiao, Taihong; Yuan, Jinwei; Adam, Hartwig; Yang, Ming–Hsuan

doi:10.1007/978-3-031-19778-9_3

Cited by 26 publications

(28 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(1) Feature-wise transformation (Chen Q. et al, 2022 ) is used to convert feature distributions learned on the source domain into that of target domain. Huang H.-P. et al ( 2022 ) introduce an integrated adaptor module and a feature transformation layer in adaptive vision transformers (ViT), which adapts to different domains with a few samples to achieve robust performance. (2) Attention mechanism can learn some essential features of the target object and improve the learning ability of the model for critical features.…”

Section: Metric-based Few-shot Image Classificationmentioning

confidence: 99%

See 1 more Smart Citation

Learning with few samples in deep learning for image classification, a mini-review

Zhang

Liu

2023

Front. Comput. Neurosci.

View full text Add to dashboard Cite

Deep learning has achieved enormous success in various computer tasks. The excellent performance depends heavily on adequate training datasets, however, it is difficult to obtain abundant samples in practical applications. Few-shot learning is proposed to address the data limitation problem in the training process, which can perform rapid learning with few samples by utilizing prior knowledge. In this paper, we focus on few-shot classification to conduct a survey about the recent methods. First, we elaborate on the definition of the few-shot classification problem. Then we propose a newly organized taxonomy, discuss the application scenarios in which each method is effective, and compare the pros and cons of different methods. We classify few-shot image classification methods from four perspectives: (i) Data augmentation, which contains sample-level and task-level data augmentation. (ii) Metric-based method, which analyzes both feature embedding and metric function. (iii) Optimization method, which is compared from the aspects of self-learning and mutual learning. (iv) Model-based method, which is discussed from the perspectives of memory-based, rapid adaptation and multi-task learning. Finally, we conduct the conclusion and prospect of this paper.

show abstract

Section: Metric-based Few-shot Image Classificationmentioning

confidence: 99%

“…is used to convert feature distributions learned on the source domain into that of target domain Huang H.-P. et al (2022). introduce an integrated adaptor module and a feature transformation layer in adaptive vision transformers (ViT), which adapts to different domains with a few samples to achieve robust performance.…”

mentioning

confidence: 99%

Learning with few samples in deep learning for image classification, a mini-review

Zhang

Liu

2023

Front. Comput. Neurosci.

View full text Add to dashboard Cite

show abstract

“…With their in-built local patchifying and global self-attention mechanisms, ViTs may be potentially better-suited to FAS over their CNN counterparts. Most recently, CNN equipped with attention modules [28] and sophisticated designed ViT variants [13,18,20,24] have been introduced into FAS and obtained promising performance. However, whether a vanilla ViT without extra training samples from upstream tasks can achieve competitive crossdomain generalization has not been explored thus far.…”

Section: Introductionmentioning

confidence: 99%

Spoofing Attacks and Anti-Spoofing Methods for Face Authentication Over Smartphones

Zhang¹,

Wang²

2023

IEEE Commun. Mag.

View full text Add to dashboard Cite

Face anti-spoofing (FAS) method performs well under the intra-domain setups. But cross-domain performance of the model is not satisfying. Domain generalization method has been used to align the feature from different domain extracted by convolutional neural network (CNN) backbone. However, the improvement is limited. Recently, the Vision Transformer (ViT) model has performed well on various visual tasks. But ViT model relies heavily on pre-training of large-scale dataset, which cannot be satisfied by existing FAS datasets. In this paper, taking the FAS task as an example, we propose Masked Contrastive Autoencoder (MCAE) method to solve this problem using only limited data. Meanwhile in order for a feature extractor to extract common features in live samples from different domains, we combine Masked Image Model (MIM) with supervised contrastive learning to train our model. Some intriguing design principles are summarized for performing MIM pre-training for downstream tasks. We also provide insightful analysis for our method from an information theory perspective. Experimental results show our approach has good performance on extensive public datasets and outperforms the state-ofthe-art methods.

show abstract

“…To tackle the cross-domain problem, domain generalization [41,13] and adaption [21,10] techniques for FAS have been extensively studied in recent years. Domain generalization-based methods aim to develop a generalized FAS model with training data from multiple source domains.…”

Section: Introductionmentioning

confidence: 99%

“…During continual sessions, a small amount of data could lead to overfitting, bring poor generalization performance and catastrophic forgetting. To update models continually and efficiently, we introduce Efficient Parameter Transfer Learning (EPTL) paragdim for the DCL-FAS and utilize Adapters [9,10] for Vision Transformer (ViT) [6]. By using the adapters, [9], ViT models can be efficiently adapted even with low-shot training data.…”

Section: Introductionmentioning

confidence: 99%

Learning Meta Pattern for Face Anti-Spoofing

Cai

Wan

et al. 2022

IEEE Trans.Inform.Forensic Secur.

View full text Add to dashboard Cite

Face Anti-Spoofing (FAS) is recently studied under the continual learning setting, where the FAS models are expected to evolve after encountering the data from new domains. However, existing methods need extra replay buffers to store previous data for rehearsal, which becomes infeasible when previous data is unavailable because of privacy issues. In this paper, we propose the first rehearsalfree method for Domain Continual Learning (DCL) of FAS, which deals with catastrophic forgetting and unseen domain generalization problems simultaneously. For better generalization to unseen domains, we design the Dynamic Central Difference Convolutional Adapter (DCDCA) to adapt Vision Transformer (ViT) models during the continual learning sessions. To alleviate the forgetting of previous domains without using previous data, we propose the Proxy Prototype Contrastive Regularization (PPCR) to constrain the continual learning with previous domain knowledge from the proxy prototypes. Simulate practical DCL scenarios, we devise two new protocols which evaluate both generalization and anti-forgetting performance. Extensive experimental results show that our proposed method can improve the generalization performance in unseen domains and alleviate the catastrophic forgetting of the previous knowledge. The codes and protocols will be released soon.

show abstract

Adaptive Transformers for Robust Few-shot Cross-domain Face Anti-spoofing

Cited by 26 publications

References 56 publications

Learning with few samples in deep learning for image classification, a mini-review

Learning with few samples in deep learning for image classification, a mini-review

Spoofing Attacks and Anti-Spoofing Methods for Face Authentication Over Smartphones

Learning Meta Pattern for Face Anti-Spoofing

Contact Info

Product

Resources

About