Face recognition under variable lighting using the mean field method and the gray-level pyramid

Son, Tran Thai; Mita, Seiichi

doi:10.1109/icsmc.2005.1571460

Cited by 3 publications

(5 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Histogram equalization can be used to illumination compensation, but, since it is based on the global pixel transforms, the correction recognition ratio with the compensation scheme is limited [2]. Xudong Xie and Kin-Man Lam presented local pixel normalization scheme.…”

Section: Local Pixel Normalizationmentioning

confidence: 99%

“…In order to avoid overflow, a small constant 0.01 is added to the variance value. From equation (2), it is obvious that after the LN processing, ) ,…”

Section: Local Pixel Normalizationmentioning

confidence: 99%

“…After local pixel normalization, opening, a morphological transformation, is utilized to reduce noises [8]. Experiment results show that the method to reduce noise is better than the method in [2].…”

Section: Local Pixel Normalizationmentioning

confidence: 99%

“…In some outdoor applications such as access control to building, lighting is often not under control. However, the variations between the images of the same face due to illumination and viewing direction are almost always larger than the image variations due to a change in face identity [1][2][3][4]. If only upright frontal face images under even illumination are available for training, face recognition based on the statistical pattern recognition such as PCA, LDA and ICA will encounter difficulties, and their performance will illumination normalization method (LN for short) to reduce effect of variable lighting on face recognition [3].The scheme utilizes the local mean and the local variance of the pixel neighbors to normalize the pixel, and utilizes a 3*3 filter to reduce noise.…”

Section: Introductionmentioning

confidence: 98%

See 3 more Smart Citations

A Half Face Recognition Scheme

Ando

2007

Eighth ACIS International Conference on Software Engineering, Artificial Intelligence, Networking, and Parallel/Distributed Com

View full text Add to dashboard Cite

A half face recognition scheme is proposed to reduce effect of variable lighting on face recognition. The scheme includes 3 stages. 1) The face images in training set are divided into the left face images and the right half face images firstly, and then they train the left feature extractor and the right feature extractor after illumination compensation respectively. 2) One half of a probe face image is chosen by entropy and its features are extracted by the corresponding feature extractor after illumination compensation. 3) The face is classified by the features of the chosen half face image. Experiment results show that, under variable lighting, the scheme not only gains the higher recognition ratio than the whole face, but also gains the higher speed.

show abstract

Section: Local Pixel Normalizationmentioning

confidence: 99%

“…In order to avoid overflow, a small constant 0.01 is added to the variance value. From equation (2), it is obvious that after the LN processing, ) ,…”

Section: Local Pixel Normalizationmentioning

confidence: 99%

Section: Local Pixel Normalizationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 98%

See 2 more Smart Citations

A Half Face Recognition Scheme

Ando

2007

Eighth ACIS International Conference on Software Engineering, Artificial Intelligence, Networking, and Parallel/Distributed Com

View full text Add to dashboard Cite

show abstract

“…We just name a few in this paragraph without intending to be complete. Son and Mita [18] adopted edge features and used the mean-field algorithm to reduce noise edges from shadow regions. A kernel-based linear discriminant analysis method, where the input space is mapped to a high-dimensional feature space, was proposed in [19] to handle large illumination variations as well as pose variations.…”

mentioning

confidence: 99%

Intrinsic Illumination Subspace for Lighting Insensitive Face Recognition

Chen

2012

IEEE Trans. Syst., Man, Cybern. B

View full text Add to dashboard Cite

Abstract-We introduce the intrinsic illumination subspace and its application for lighting insensitive face recognition in this paper. The intrinsic illumination subspace is constructed from illumination images of intrinsic images, which is a midlevel description of appearance images and can be useful for many visual inferences. This subspace forms a convex polyhedral cone and can be efficiently represented by a low-dimensional linear subspace, which enables an analytic generation of illumination images under varying lighting conditions. When only objects of the same class, such as faces, are concerned, a class-based generic intrinsic illumination subspace can be constructed in advance and used for novel objects of the same class. Based on this class-based generic subspace, we propose a lighting normalization method for lighting insensitive face recognition, where only a single input image is required. The generic subspace is used as a bootstrap subspace for illumination images of novel objects. Face recognition experiments are performed to demonstrate the effectiveness of the proposed lighting normalization method and verify empirically that the class-based generic subspace is applicable to novel objects. Our method is simple and fast, which makes it useful for real-time applications, embedded systems, or mobile devices with limited resources.

show abstract

An Attitude-adaptation Negotiation Strategy in Electronic Market Environments

Dong

Jiang

Wang

et al. 2007

Eighth ACIS International Conference on Software Engineering, Artificial Intelligence, Networking, and Parallel/Distributed Com

View full text Add to dashboard Cite

The automatization of electronic commerce negotiation has become the focus of more and more attention. As one of the core of automated negotiation, strategy is the method employed by agents to maximize their own benefits. The design of negotiation strategy is affected by lots of factors, such as negotiation deadline, the type of resources, the characteristics of opponents, the numbers of competitors, etc. Though lots of work have been done on negotiation strategy, there still lack a uniform strategy framework. In this paper, a general description framework for multiagent negotiation is proposed. Then solutions to the decision models are given and the existence of the Nash equilibrium in agents' attitude selection at the beginning of negotiations is analyzed. To enable agents to adapt their attitudes according to negotiation duration, the supply and demand ratio, and the characteristics of opponent, we introduce an adaptation function. This function is constructed on the basis of a history-based learning algorithm. Furthermore, the attitude-adaptation strategy and the fixed-attitude strategies are compared by simulation.

show abstract

Face recognition under variable lighting using the mean field method and the gray-level pyramid

Cited by 3 publications

References 6 publications

A Half Face Recognition Scheme

A Half Face Recognition Scheme

Intrinsic Illumination Subspace for Lighting Insensitive Face Recognition

An Attitude-adaptation Negotiation Strategy in Electronic Market Environments

Contact Info

Product

Resources

About