# Robust LSTM-Autoencoders for Face De-Occlusion in the Wild

arXiv: https://arxiv.org/abs/1612.08534 · HTML: https://mlsanity.com/paper/1612.08534/

## Key facts

- Cited by 17 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Dec 2016.

## Built on by (most cited first, 17 of 17)

- [End2End Occluded Face Recognition by Masking Corrupted Features](https://mlsanity.com/paper/2108.09468/) (2021)
- [DeepFace-EMD: Re-ranking Using Patch-wise Earth Mover's Distance Improves Out-Of-Distribution Face Identification](https://mlsanity.com/paper/2112.04016/) (2021)
- [Mask-FPAN: Semi-Supervised Face Parsing in the Wild With De-Occlusion and UV GAN](https://mlsanity.com/paper/2212.09098/) (2022)
- [Revisiting Bayesian Autoencoders with MCMC](https://mlsanity.com/paper/2104.05915/) (2021)
- [Segmentation-Reconstruction-Guided Facial Image De-occlusion](https://mlsanity.com/paper/2112.08022/) (2021)
- [Fun Selfie Filters in Face Recognition: Impact Assessment and Removal](https://mlsanity.com/paper/2202.06022/) (2022)
- [Depth as Attention for Face Representation Learning](https://mlsanity.com/paper/2101.00652/) (2021)
- [Locality-aware Channel-wise Dropout for Occluded Face Recognition](https://mlsanity.com/paper/2107.09270/) (2021)
- [Fast and Interpretable Face Identification for Out-Of-Distribution Data Using Vision Transformers](https://mlsanity.com/paper/2311.02803/) (2023)
- [A Unified Framework for Masked and Mask-Free Face Recognition via Feature Rectification](https://mlsanity.com/paper/2202.07358/) (2022)
- [Dominant motion identification of multi-particle system using deep learning from video](https://mlsanity.com/paper/2104.12722/) (2021)
- [Towards robustness under occlusion for face recognition](https://mlsanity.com/paper/2109.09083/) (2021)
- [A Survey of Face Recognition](https://mlsanity.com/paper/2212.13038/) (2022)
- [State of the Art: Face Recognition](https://mlsanity.com/paper/2108.11821/) (2021)
- [Learning Representations for Masked Facial Recovery](https://mlsanity.com/paper/2212.14110/) (2022)
- [Seeing through the Mask: Multi-task Generative Mask Decoupling Face Recognition](https://mlsanity.com/paper/2311.11512/) (2023)
- [Geometric Analysis of Magnetic Labyrinthine Stripe Evolution via Deep Learning Segmentation](https://mlsanity.com/paper/2509.11485/) (2025)

## Abstract (arXiv)

Face recognition techniques have been developed significantly in recent years. However, recognizing faces with partial occlusion is still challenging for existing face recognizers which is heavily desired in real-world applications concerning surveillance and security. Although much research effort has been devoted to developing face de-occlusion methods, most of them can only work well under constrained conditions, such as all the faces are from a pre-defined closed set. In this paper, we propose a robust LSTM-Autoencoders (RLA) model to effectively restore partially occluded faces even in the wild. The RLA model consists of two LSTM components, which aims at occlusion-robust face encoding and recurrent occlusion removal respectively. The first one, named multi-scale spatial LSTM encoder, reads facial patches of various scales sequentially to output a latent representation, and occlusion-robustness is achieved owing to the fact that the influence of occlusion is only upon some of the patches. Receiving the representation learned by the encoder, the LSTM decoder with a dual channel architecture reconstructs the overall face and detects occlusion simultaneously, and by feat of LSTM, the decoder breaks down the task of face de-occlusion into restoring the occluded part step by step. Moreover, to minimize identify information loss and guarantee face recognition accuracy over recovered faces, we introduce an identity-preserving adversarial training scheme to further improve RLA. Extensive experiments on both synthetic and real datasets of faces with occlusion clearly demonstrate the effectiveness of our proposed RLA in removing different types of facial occlusion at various locations. The proposed method also provides significantly larger performance gain than other de-occlusion methods in promoting recognition performance over partially-occluded faces.