Latent Factor Guided Convolutional Neural Networks for Age-Invariant Face Recognition

Wen, Yandong; Li, Zhifeng; Qiao, Yu

doi:10.1109/cvpr.2016.529

Cited by 144 publications

(120 citation statements)

References 23 publications

Supporting

Mentioning

117

Contrasting

Unclassified

Order By: Relevance

“…For example , the [13] separates the identity-related information and the age-related information through the hidden factor analysis (HFA). The [48] is based on similar analysis and extends the HFA to the deep learning framework. More recently, the OE-CNN [46] presents the orthogonal feature decomposition to solve the AIFR.…”

Section: Intra-identity Distancementioning

confidence: 99%

Decorrelated Adversarial Learning for Age-Invariant Face Recognition

Wang

Gong

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Self Cite

View full text Add to dashboard Cite

There has been an increasing research interest in ageinvariant face recognition. However, matching faces with big age gaps remains a challenging problem, primarily due to the significant discrepancy of face appearances caused by aging. To reduce such a discrepancy, in this paper we propose a novel algorithm to remove age-related components from features mixed with both identity and age information. Specifically, we factorize a mixed face feature into two uncorrelated components: identity-dependent component and age-dependent component, where the identitydependent component includes information that is useful for face recognition. To implement this idea, we propose the Decorrelated Adversarial Learning (DAL) algorithm, where a Canonical Mapping Module (CMM) is introduced to find the maximum correlation between the paired features generated by a backbone network, while the backbone network and the factorization module are trained to generate features reducing the correlation. Thus, the proposed model learns the decomposed features of age and identity whose correlation is significantly reduced. Simultaneously, the identity-dependent feature and the age-dependent feature are respectively supervised by ID and age preserving signals to ensure that they both contain the correct information. Extensive experiments are conducted on popular public-domain face aging datasets (FG-NET, MORPH Album 2, and CACD-VS) to demonstrate the effectiveness of the proposed approach.

show abstract

Section: Intra-identity Distancementioning

confidence: 99%

Decorrelated Adversarial Learning for Age-Invariant Face Recognition

Wang

Gong

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Alternatively we can finetune the model from the pre-trained face recognition model., However, it may not be helpful for all the models and we found directly finetuning may fall into local extreme. To solve the training difficulty problem, we adopted the mixed loss [13],the formula is as follows…”

Section: 3improved Loss Ensemblementioning

confidence: 99%

Video Age Estimation with Multiple Stacked CNN Models

Zhang¹

2017

Proceedings of the 7th International Conference on Computer Engineering and Networks — PoS(CENet2017)

View full text Add to dashboard Cite

Automatic age classification has become relevant to an increasing amount of applications, particularly after the occurence of many social platforms and social medias where the video age recognition is important for the improvement of user experience; however, performance of the existing methods on real-world video continuous images is in great shortage, especially when it is compared to the "superhuman" improvement of recognition precision reported for the related task of object and face recognition. In this paper, we proposed a new cnn structure by combining several stacked deep-convolutional neural networks (CNN), which consists of an improved alexnet and an improved grouped googlenet. The stacked models can be used to estimate the apparent age of the people from coarse to fine. Experiments showed that a significant improvement in performance can be obtained on the video tasks. We evaluated our method on the recent benchmark for video apparent age estimation and showed it to outperform current state-of-the-art methods.

show abstract

“…Subsequently, a joint identificationverification supervision signal was adopted in Refs. [10,13], leading to more discriminative representation features. Reference [16] enhanced supervision by adding a fully connected layer and loss functions to each convolutional layer.…”

Section: Introductionmentioning

confidence: 99%

“…Wang et al [6] proposed a multitask learning structure using local binary patterns (LBP) [7] to solve face verification and retrieval problems. Learning face representations via deep learning has achieved a series of breakthroughs in recent years [8][9][10][11][12][13]. The idea of mapping a pair of face images to a distance originated in Ref.…”

Section: Introductionmentioning

confidence: 99%

Face image retrieval based on shape and texture feature fusion

Yang

Liu

2017

Comp. Visual Media

View full text Add to dashboard Cite

Humongous amounts of data bring various challenges to face image retrieval. This paper proposes an efficient method to solve those problems. Firstly, we use accurate facial landmark locations as shape features. Secondly, we utilise shape priors to provide discriminative texture features for convolutional neural networks. These shape and texture features are fused to make the learned representation more robust. Finally, in order to increase efficiency, a coarse-tofine search mechanism is exploited to efficiently find similar objects. Extensive experiments on the CASIAWebFace, MSRA-CFW, and LFW datasets illustrate the superiority of our method.

show abstract

Latent Factor Guided Convolutional Neural Networks for Age-Invariant Face Recognition

Cited by 144 publications

References 23 publications

Decorrelated Adversarial Learning for Age-Invariant Face Recognition

Decorrelated Adversarial Learning for Age-Invariant Face Recognition

Video Age Estimation with Multiple Stacked CNN Models

Face image retrieval based on shape and texture feature fusion

Contact Info

Product

Resources

About