# Learnable Multi-level Frequency Decomposition and Hierarchical Attention Mechanism for Generalized Face Presentation Attack Detection

arXiv: https://arxiv.org/abs/2109.07950 · HTML: https://mlsanity.com/paper/2109.07950/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- Builds on 15 of its 36 references that are in the map.
- First posted on arXiv 16 Sep 2021.

## Builds on (most cited first, 15 of 15)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [CBAM: Convolutional Block Attention Module](https://mlsanity.com/paper/1807.06521/) (2018)
- [Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks](https://mlsanity.com/paper/1604.02878/) (2016)
- [Thinking in Frequency: Face Forgery Detection by Mining Frequency-aware Clues](https://mlsanity.com/paper/2007.09355/) (2020)
- [Searching Central Difference Convolutional Networks for Face Anti-Spoofing](https://mlsanity.com/paper/2003.04092/) (2020)
- [Learning Deep Models for Face Anti-Spoofing: Binary or Auxiliary Supervision](https://mlsanity.com/paper/1803.11097/) (2018)
- [NAS-FAS: Static-Dynamic Central Difference Network Search for Face Anti-Spoofing](https://mlsanity.com/paper/2011.02062/) (2020)
- [Regularized Fine-grained Meta Face Anti-spoofing](https://mlsanity.com/paper/1911.10771/) (2019)
- [Deep Pixel-wise Binary Supervision for Face Presentation Attack Detection](https://mlsanity.com/paper/1907.04047/) (2019)
- [Face De-Spoofing: Anti-Spoofing via Noise Modeling](https://mlsanity.com/paper/1807.09968/) (2018)
- [Domain Agnostic Feature Learning for Image and Video Based Face Anti-spoofing](https://mlsanity.com/paper/1912.07124/) (2019)
- [Look Locally Infer Globally: A Generalizable Face Anti-Spoofing Approach](https://mlsanity.com/paper/2006.02834/) (2020)
- [Iris Presentation Attack Detection by Attention-based and Deep Pixel-wise Binary Supervision Network](https://mlsanity.com/paper/2106.14845/) (2021)
- [Exploiting temporal and depth information for multi-frame face anti-spoofing](https://mlsanity.com/paper/1811.05118/) (2018)
- [Camera Invariant Feature Learning for Generalized Face Anti-spoofing](https://mlsanity.com/paper/2101.10075/) (2021)

## Built on by (most cited first, 18 of 18)

- [SynthASpoof: Developing Face Presentation Attack Detection Based on Privacy-friendly Synthetic Data](https://mlsanity.com/paper/2303.02660/) (2023)
- [FoundPAD: Foundation Models Reloaded for Face Presentation Attack Detection](https://mlsanity.com/paper/2501.02892/) (2025)
- [Fairness in Face Presentation Attack Detection](https://mlsanity.com/paper/2209.09035/) (2022)
- [Partial Attack Supervision and Regional Weighted Inference for Masked Face Presentation Attack Detection](https://mlsanity.com/paper/2111.04336/) (2021)
- [Face Presentation Attack Detection by Excavating Causal Clues and Adapting Embedding Statistics](https://mlsanity.com/paper/2308.14551/) (2023)
- [Are Explainability Tools Gender Biased? A Case Study on Face Presentation Attack Detection](https://mlsanity.com/paper/2304.13419/) (2023)
- [Liveness score-based regression neural networks for face anti-spoofing](https://mlsanity.com/paper/2302.09461/) (2023)
- [A Survey on Computer Vision based Human Analysis in the COVID-19 Era](https://mlsanity.com/paper/2211.03705/) (2022)
- [Few-shot One-class Domain Adaptation Based on Frequency for Iris Presentation Attack Detection](https://mlsanity.com/paper/2204.00376/) (2022)
- [Intra and Cross-spectrum Iris Presentation Attack Detection in the NIR and Visible Domains](https://mlsanity.com/paper/2205.02573/) (2022)
- [Flickr-PAD: New Face High-Resolution Presentation Attack Detection Database](https://mlsanity.com/paper/2304.13015/) (2023)
- [A Closer Look at Geometric Temporal Dynamics for Face Anti-Spoofing](https://mlsanity.com/paper/2306.14313/) (2023)
- [PAD-Phys: Exploiting Physiology for Presentation Attack Detection in Face Biometrics](https://mlsanity.com/paper/2310.02140/) (2023)
- [Foundation Models for Face Presentation Attack Detection: A Unified Linear-Probing Benchmark](https://mlsanity.com/paper/2607.26993/) (2026)
- [In-context Learning of Vision Language Models for Detection of Physical and Digital Attacks against Face Recognition Systems](https://mlsanity.com/paper/2507.15285/) (2025)
- [Introducing Nylon Face Mask Attacks: A Dataset for Evaluating Generalised Face Presentation Attack Detection](https://mlsanity.com/paper/2511.08114/) (2025)
- [LoRA-based Adaptation Alone Is Not Enough: Understanding the Limits of Foundation Models for Face Presentation Attack Detection](https://mlsanity.com/paper/2608.09633/) (2026)
- [Foundation and Multimodal Large Language Models for Face Presentation and Morph Attack Detection](https://mlsanity.com/paper/2608.29802/) (2026)

## Abstract (arXiv)

With the increased deployment of face recognition systems in our daily lives, face presentation attack detection (PAD) is attracting much attention and playing a key role in securing face recognition systems. Despite the great performance achieved by the hand-crafted and deep-learning-based methods in intra-dataset evaluations, the performance drops when dealing with unseen scenarios. In this work, we propose a dual-stream convolution neural networks (CNNs) framework. One stream adapts four learnable frequency filters to learn features in the frequency domain, which are less influenced by variations in sensors/illuminations. The other stream leverages the RGB images to complement the features of the frequency domain. Moreover, we propose a hierarchical attention module integration to join the information from the two streams at different stages by considering the nature of deep features in different layers of the CNN. The proposed method is evaluated in the intra-dataset and cross-dataset setups, and the results demonstrate that our proposed approach enhances the generalizability in most experimental setups in comparison to state-of-the-art, including the methods designed explicitly for domain adaption/shift problems. We successfully prove the design of our proposed PAD solution in a step-wise ablation study that involves our proposed learnable frequency decomposition, our hierarchical attention module design, and the used loss function. Training codes and pre-trained models are publicly released