LDF-Net: Learning a Displacement Field Network for Face Recognition across Pose

Hu, Lanqing; Kan, Meina; Shan, Shiguang; Song, Xianfeng; Chen, Xilin

doi:10.1109/fg.2017.12

Cited by 14 publications

(7 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They devised a novel network, namely DR-GAN, which simultaneously synthesizes frontal faces and learn pose-invariant feature representations. Hu et al [52] proposed to directly transform a non-frontal face into frontal face by Learning a Displacement Field network (LDF-Net). LDF-Net achieves state-of-the-art performance for face recognition across poses on Multi-PIE, especially at large poses.…”

Section: Unconstrained Face Recognitionmentioning

confidence: 99%

“…Here, we consider pose and expression variations, and conduct two experiments. In the first experiment, following the setting of [3], [75], probe images consist of the images of all 337 [78] and LDF-Net [52]. The first three matchers are publicly available.…”

Section: Face Identification On Multi-pie Databasementioning

confidence: 99%

“…In the second experiment, instead of neutral expression, all images with smile, surprise, squint, disgust and scream expressions at the 12 poses and under frontal illumination are the probe images. This protocol is an extended version of [3], [4] by adding large-pose images (±60 Lightened CNN [77], CenterLoss [78] and LDF-Net [52]. The first three matchers are publicly available.…”

Section: Face Identification On Multi-pie Databasementioning

confidence: 99%

See 2 more Smart Citations

Joint Face Alignment and 3D Face Reconstruction with Application to Face Recognition

Liu

Zhao

Liu

et al. 2020

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

Face alignment and 3D face reconstruction are traditionally accomplished as separated tasks. By exploring the strong correlation between 2D landmarks and 3D shapes, in contrast, we propose a joint face alignment and 3D face reconstruction method to simultaneously solve these two problems for 2D face images of arbitrary poses and expressions. This method, based on a summation model of 3D faces and cascaded regression in 2D and 3D shape spaces, iteratively and alternately applies two cascaded regressors, one for updating 2D landmarks and the other for 3D shape. The 3D shape and the landmarks are correlated via a 3D-to-2D mapping matrix, which is updated in each iteration to refine the location and visibility of 2D landmarks. Unlike existing methods, the proposed method can fully automatically generate both pose-and-expression-normalized (PEN) and expressive 3D faces and localize both visible and invisible 2D landmarks. Based on the PEN 3D faces, we devise a method to enhance face recognition accuracy across poses and expressions. Both linear and nonlinear implementations of the proposed method are presented and evaluated in this paper. Extensive experiments show that the proposed method can achieve the state-of-the-art accuracy in both face alignment and 3D face reconstruction, and benefit face recognition owing to its reconstructed PEN 3D face.

show abstract

Section: Unconstrained Face Recognitionmentioning

confidence: 99%

Section: Face Identification On Multi-pie Databasementioning

confidence: 99%

Section: Face Identification On Multi-pie Databasementioning

confidence: 99%

See 1 more Smart Citation

Joint Face Alignment and 3D Face Reconstruction with Application to Face Recognition

Liu

Zhao

Liu

et al. 2020

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

show abstract

“…Brief Description Subsettings one to many generating many patches or images of the pose variability from a single image 3D model [139], [137], [165], [166], [53] [67], [197], [196] 2D deep model [279], [267], [182] data augmentation [124], [276], [51] [222], [187], [188], [192], [202] many to one recovering the canonical view of face images from one or many images of nonfrontal view SAE [101], [264], [240] CNN [278], [280], [89], [37], [246] GAN [91], [198], [41], [249]…”

Section: Data Processingmentioning

confidence: 99%

“…Yim et al [246] proposed a multi-task network that can rotate an arbitrary pose and illumination image to the target-pose face image by utilizing the user's remote code. [89] transformed nonfrontal face images to frontal images according to the displacement field of the pixels between them. Zhou et al [275] proposed a novel non-rigid face rectification method by local homography transformations, and regularized it by imposing natural frontal face distribution with a denoising autoencoder.…”

Section: B Many-to-one Normalizationmentioning

confidence: 99%

Deep Face Recognition: A Survey

Wang,

Deng

2018

Preprint

View full text Add to dashboard Cite

Deep learning applies multiple processing layers to learn representations of data with multiple levels of feature extraction. This emerging technique has reshaped the research landscape of face recognition (FR) since 2014, launched by the breakthroughs of Deepface method. Since then, deep FR technique, which leverages hierarchical architecture to stitch together pixels into invariant face representation, has dramatically improved the state-of-the-art performance and fostered successful real-world applications. In this survey, we provide a comprehensive review of the recent developments on deep FR, covering both broad topics on algorithm designs, databases and protocols, and application scenes. First, we summarize different network architectures and loss functions proposed in the rapid evolution of the deep FR methods. Second, the related face processing methods are categorized into two classes: "one-to-many augmentation" and "many-to-one normalization". Then, we summarize and compare the commonly used databases for both model training and evaluation. Third, we review miscellaneous scenes in deep FR, such as cross-factor, heterogenous, multiplemedia and industry scenes. Finally, the technical challenges of current methods and several promising directions on biometric security, fusion, and privacy are highlighted.

show abstract