# PFLD: A Practical Facial Landmark Detector

arXiv: https://arxiv.org/abs/1902.10859 · HTML: https://mlsanity.com/paper/1902.10859/

## Key facts

- Cited by 26 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 28 Feb 2019.

## Built on by (most cited first, 26 of 26)

- [MetaPortrait: Identity-Preserving Talking Head Generation with Fast Personalized Adaptation](https://mlsanity.com/paper/2212.08062/) (2022)
- [Implicit Identity Representation Conditioned Memory Compensation Network for Talking Head video Generation](https://mlsanity.com/paper/2307.09906/) (2023)
- [FaceX-Zoo: A PyTorch Toolbox for Face Recognition](https://mlsanity.com/paper/2101.04407/) (2021)
- [From Static to Dynamic: Adapting Landmark-Aware Image Models for Facial Expression Recognition in Videos](https://mlsanity.com/paper/2312.05447/) (2023)
- [Py-Feat: Python Facial Expression Analysis Toolbox](https://mlsanity.com/paper/2104.03509/) (2021)
- [Personalized Image Generation with Deep Generative Models: A Decade Survey](https://mlsanity.com/paper/2502.13081/) (2025)
- [Fast Facial Landmark Detection and Applications: A Survey](https://mlsanity.com/paper/2101.10808/) (2021)
- [DuetFace: Collaborative Privacy-Preserving Face Recognition via Channel Splitting in the Frequency Domain](https://mlsanity.com/paper/2207.07340/) (2022)
- [3DCaricShop: A Dataset and A Baseline Method for Single-view 3D Caricature Face Reconstruction](https://mlsanity.com/paper/2103.08204/) (2021)
- [Mask-FPAN: Semi-Supervised Face Parsing in the Wild With De-Occlusion and UV GAN](https://mlsanity.com/paper/2212.09098/) (2022)
- [Facial Landmark Points Detection Using Knowledge Distillation-Based Neural Networks](https://mlsanity.com/paper/2111.07047/) (2021)
- [VQualA 2025 Challenge on Face Image Quality Assessment: Methods and Results](https://mlsanity.com/paper/2508.18445/) (2025)
- [Real-time Driver Monitoring Systems on Edge AI Device](https://mlsanity.com/paper/2304.01555/) (2023)
- [ORFormer: Occlusion-Robust Transformer for Accurate Facial Landmark Detection](https://mlsanity.com/paper/2412.13174/) (2024)
- [DogFLW: Dog Facial Landmarks in the Wild Dataset](https://mlsanity.com/paper/2405.11501/) (2024)
- [Automated Assessment of Aesthetic Outcomes in Facial Plastic Surgery](https://mlsanity.com/paper/2508.13363/) (2025)
- [Fast Eye Detector Using Siamese Network for NIR Partial Face Images](https://mlsanity.com/paper/2202.10671/) (2022)
- [A Socially Assistive Robot using Automated Planning in a Paediatric Clinical Setting](https://mlsanity.com/paper/2210.09753/) (2022)
- [A Comparative Analysis of the Face Recognition Methods in Video Surveillance Scenarios](https://mlsanity.com/paper/2211.02952/) (2022)
- [Vision-based Estimation of Fatigue and Engagement in Cognitive Training Sessions](https://mlsanity.com/paper/2304.12470/) (2023)
- [Superior and Pragmatic Talking Face Generation with Teacher-Student Framework](https://mlsanity.com/paper/2403.17883/) (2024)
- [Deep Learning-Based Quasi-Conformal Surface Registration for Partial 3D Faces Applied to Facial Recognition](https://mlsanity.com/paper/2405.09880/) (2024)
- [Efficient Facial Landmark Detection for Embedded Systems](https://mlsanity.com/paper/2407.10228/) (2024)
- [Audio-Assisted Face Video Restoration with Temporal and Identity Complementary Learning](https://mlsanity.com/paper/2508.04161/) (2025)
- [Facial Expression Recognition in the Deep Learning Era: A Systematic Multi-Criteria Review of Methods, Models, Datasets, Performance, Challenges, and Future Research Directions](https://mlsanity.com/paper/2606.08612/) (2026)
- [NeoLoc-68: End-to-end 68-point neonatal facial landmark localisation in neonatal clinical environments](https://mlsanity.com/paper/2606.20823/) (2026)

## Abstract (arXiv)

Being accurate, efficient, and compact is essential to a facial landmark detector for practical use. To simultaneously consider the three concerns, this paper investigates a neat model with promising detection accuracy under wild environments e.g., unconstrained pose, expression, lighting, and occlusion conditions) and super real-time speed on a mobile device. More concretely, we customize an end-to-end single stage network associated with acceleration techniques. During the training phase, for each sample, rotation information is estimated for geometrically regularizing landmark localization, which is then NOT involved in the testing phase. A novel loss is designed to, besides considering the geometrical regularization, mitigate the issue of data imbalance by adjusting weights of samples to different states, such as large pose, extreme lighting, and occlusion, in the training set. Extensive experiments are conducted to demonstrate the efficacy of our design and reveal its superior performance over state-of-the-art alternatives on widely-adopted challenging benchmarks, i.e., 300W (including iBUG, LFPW, AFW, HELEN, and XM2VTS) and AFLW. Our model can be merely 2.1Mb of size and reach over 140 fps per face on a mobile phone (Qualcomm ARM 845 processor) with high precision, making it attractive for large-scale or real-time applications. We have made our practical system based on PFLD 0.25X model publicly available at urlhttp://sites.google.com/view/xjguo/fld for encouraging comparisons and improvements from the community.