Aging Face Recognition: A Hierarchical Learning Model Based on Local Patterns Selection

ABSTRACT:

Aging face recognition refers to matching the same person’s faces across different ages, e.g., matching a person’s older face to his (or her) younger one, which has many important practical applications, such as finding missing children. The major challenge of this task is that facial appearance is subject to significant change during the aging process. In this paper, we propose to solve the problem with a hierarchical model based on two-level learning. At the first level, effective features are learned from low-level microstructures, based on our new feature descriptor called local pattern selection (LPS). The proposed LPS descriptor greedily selects low-level discriminant patterns in a way, such that intra-user dissimilarity is minimized. At the second level, higher level visual information is further refined based on the output from the first level. To evaluate the performance of our new method, we conduct extensive experiments on the MORPH data set (the largest face aging data set available in the public domain), which show a significant improvement in accuracy over the state-of-the-art methods.

PROJECT OUTPUT VIDEO:

EXISTING SYSTEM:

Age Estimation from facial images has been receiving increasing interest due to its important applications. Among the existing age estimation algorithms, the personalized approaches have been shown to be the most effective ones. However, most of the person-specific approaches (e.g. MTWGP [1], AGES [2], WAS [3]) rely heavily on the availability of training images across different ages for a single subject.

DISADVANTAGES OF EXISTING SYSTEM:

• The existing system is very difficult to satisfy in practical applications.

• Due to the strong parametric assumptions and the complexity of the algorithm, these methods are expensive to compute and the results are often unstable for real world face recognition.

PROPOSED SYSTEM:

• In this paper, we propose a two-level hierarchical learning model .In this model, effective features are first learned from low-level pixel structures, based on our new feature extraction algorithm called Local Pattern Selection (LPS). Low-level common information is widely believed to be very beneficial to cross-age face recognition, and the LPS algorithm maximizes this information between cross-age faces.

• At the second level, higher-level visual information is refined by learning subspace analysis algorithms. The advantage of this model is that, when com-pared with traditional paradigms where learning happens only in higher levels via classification algorithms, our model has better learning capabilities and is thus able to adaptively capture more useful information. Also note that strong model learning capability is essential to a successful face recognition algorithm, as confirmed by the recent success in deep learning. Extensive experiments are conducted on the MORPH dataset (Album 2), the largest publicly available facial aging dataset, to validate the effectiveness of our new approach over the state-of-the-art ones.

ADVANTAGES OF PROPOSED SYSTEM:

• Our system is used to enhance the age estimation performance.

• The advantage of this model is that, when compared with traditional paradigms where learning happens only in higher levels via classification algorithms, our model has better learning capabilities and is thus able to adaptively capture more useful information.

• Also note that strong model learning capability is essential to a successful face recognition algorithm, as confirmed by the recent success in deep learning

MODULES:

• Image Acquisition
• Preprocessing
• Feature Extraction
• Refinement
• Face Recognition

MODULE DESCRIPTION:

Image Acquisition:

• The image has been obtained by Gallery.

Preprocessing:

Preprocessing having two steps,

1. Face Detection
2. Sampling

i. Face Detection:

After acquire the image from gallery, detect face using face detection tool. Then detected face image cropped for further process.

ii. Sampling:

In this stage cropped face image first converted into gray image then sampling method is performed. Also histogram was generated for gray image.

The sampling rate determines the spatial resolution of the digitized image. After that,pixel distribution calculation in the form of histogram.

Feature Extrcation:

In feature extraction stage, sampled face image are performed. In this stage Local Pattern Selection and Local Binary Pattern are used for feature extraction. High level visual features are extracted.

• By using LPS, we extract over-completed features based on encoded images. Following steps are involved in LPS,

1. First sampling the image with different radius.

2. After construct the encoding tree using Quadtree Decomposition. Finally get encoded image for further process.

3. Then extract the local visual feature from encoded image.

4. LPS is used for extract the common information between cross age faces.

• By using LBP, we extract the visual feature. LBP is a type of visual descriptor. Following steps are involved in LBP,

1. Divide the examined window into cells (e.g. 16×16 pixels for each cell).

2. For each pixel in a cell, compare the pixel to each of its 8 neighbors (on its left-top, left-middle, left-bottom, right-top, etc.). Follow the pixels along a circle, i.e. clockwise or counter-clockwise.

3. Where the center pixel’s value is greater than the neighbor’s value, write “1”. Otherwise, write “0”. This gives an 8-digit binary number (which is usually converted to decimal for convenience).

4. Compute the histogram, over the cell, of the frequency of each “number” occurring (i.e., each combination of which pixels are smaller and which are greater than the center).

5. Optionally normalize the histogram.

6. Concatenate (normalized) histograms of all cells. This gives the feature vector for the window.

Finally we are getting features from LPS and LBP.

Refinement:

In this stage, extracted features are refined. Because LPS having high dimensionality features that means there might be a lot of redundant information as well as noise among these features. PCA is applied for refinement process. After apply the PCA for LPS and LBP, we are getting low level visual features.

Face Recognition:

After refinement process, we extract the final features such as mean, variance, entropy, standard deviation and contrast from the refined visual features. Then these features construct a feature vector.

Finally face recognized by find distance between test feature vector and trained feature vector.

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS:

• System            : Pentium Dual Core.

• Hard Disk       : 120 GB.

• Monitor          : 15’’ LED

• Input Devices : Keyboard, Mouse

• Ram                : 1GB.

SOFTWARE REQUIREMENTS:

• Operating system : Windows 7.

• Coding Language : MATLAB

• Tool                       : MATLAB R2013A

REFERENCE:

Zhifeng Li, Senior Member, IEEE, Dihong Gong, Xuelong Li, Fellow, IEEE, and Dacheng Tao, Fellow, IEEE, “Aging Face Recognition: A Hierarchical Learning Model Based on Local Patterns Selection”, IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 25, NO. 5, MAY 2016.