Active Appearance Model Model Research Articles

Replicating Neuroscience Observations on ML/MF and AM Face Patches by Deep Generative Model.

A recent Cell paper (Chang & Tsao, 2017) reports an interesting discovery. For the face stimuli generated by a pretrained active appearance model (AAM), the responses of neurons in the areas of the primate brain that are responsible for face recognition exhibit a strong linear relationship with the shape variables and appearance variables of the AAM that generates the face stimuli. In this letter, we show that this behavior can be replicated by a deep generative model, the generator network, that assumes that the observed signals are generated by latent random variables via a top-down convolutional neural network. Specifically, we learn the generator network from the face images generated by a pretrained AAM model using a variational autoencoder, and we show that the inferred latent variables of the learned generator network have a strong linear relationship with the shape and appearance variables of the AAM model that generates the face images. Unlike the AAM model, which has an explicit shape model where the shape variables generate the control points or landmarks, the generator network has no such shape model and shape variables. Yet it can learn the shape knowledge in the sense that some of the latent variables of the learned generator network capture the shape variations in the face images generated by AAM.

Active Appearance Model for Age Prediction: A Comparison

Individual age gives key demographic data. It is viewed as a paramount delicate biometric characteristic for individual identification, contrasted with other pattern recognition issues. Age estimation is a complex issue particularly in relation to facial pictures with different ages, since the aging procedure varies extraordinarily across different age groups. In this work, we investigate deep learning techniques for age prediction based on Active Appearance Models (AAM) and six classifiers: Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Support Vector Regression (SVR), Canonical Correlation Analysis (CCA), Linear Discriminant Analysis (LDA) and Projection Twin Support Vector Machine (PTSVM) to improve the precision of age prediction based on the present methods. In this algorithm, we extracted the traits of the facial images as traits vectors using AAM model, and the classifiers are utilized to predict the age. We were able to recognize that the accuracy of CCA algorithm is the best, the intermediate is SVR and the KNN algorithm is the lowest.

Shape-appearance-correlated active appearance model

Among the challenges faced by current active shape or appearance models, facial-feature localization in the wild, with occlusion in a novel face image, i.e. in a generic environment, is regarded as one of the most difficult computer-vision tasks. In this paper, we propose an Active Appearance Model (AAM) to tackle the problem of generic environment. Firstly, a fast face-model initialization scheme is proposed, based on the idea that the local appearance of feature points can be accurately approximated with locality constraints. Nearest neighbors, which have similar poses and textures to a test face, are retrieved from a training set for constructing the initial face model. To further improve the fitting of the initial model to the test face, an orthogonal CCA (oCCA) is employed to increase the correlation between shape features and appearance features represented by Principal Component Analysis (PCA). With these two contributions, we propose a novel AAM, namely the shape-appearance-correlated AAM (SAC-AAM), and the optimization is solved by using the recently proposed fast simultaneous inverse compositional (Fast-SIC) algorithm. Experiment results demonstrate a 5–10% improvement on controlled and semi-controlled datasets, and with around 10% improvement on wild face datasets in terms of fitting accuracy compared to other state-of-the-art AAM models.

Feasibility Study of a Distributed and Parallel Environment for Implementing the Standard Version of AAM Model

The Active Appearance Model (AAM) is a class of deformable models, which, in the segmentation process, integrates the priori knowledge on the shape and the texture and deformation of the structures studied. This model in its sequential form is computationally intensive and operates on large data sets. This paper presents another framework to implement the standard version of the AAM model. We suggest a distributed and parallel approach justified by the characteristics of the model and their potentialities. We introduce a schema for the representation of the overall model and we study of operations that can be parallelized. This approach is intended to exploit the benefits build in the area of advanced image processing.

Active Appearance Model Algorithm with K-Nearest Neighbor Classifier for Face Pose Estimation

In this paper, a face pose estimation (FPE) algorithm using active appearance model (AAM) with a k-nearest neighbor (kNN) classifier is presented. AAM is a model-based image representation method used to describe non-rigid visual objects, with both shape and texture variations, using a mean vector and linear combinations of a set of variation modes. Since AAM is a deformable model, it has several variations. Owing to the variations, the model is adjusted to the input test face image using iterative searching and fitting. The error, which measures the difference between the model and a test image, is minimized with the proposed searching algorithm. The face pose is then estimated using the distances between the landmark points in the AAM model with a kNN classifier. Experimental results demonstrate that the proposed FPE algorithm can fit the face location with different face poses, with or without a hat, even wearing glasses, and identify the face pose accurately.

Efficient parallel implementation of active appearance model fitting algorithm on GPU.

The active appearance model (AAM) is one of the most powerful model-based object detecting and tracking methods which has been widely used in various situations. However, the high-dimensional texture representation causes very time-consuming computations, which makes the AAM difficult to apply to real-time systems. The emergence of modern graphics processing units (GPUs) that feature a many-core, fine-grained parallel architecture provides new and promising solutions to overcome the computational challenge. In this paper, we propose an efficient parallel implementation of the AAM fitting algorithm on GPUs. Our design idea is fine grain parallelism in which we distribute the texture data of the AAM, in pixels, to thousands of parallel GPU threads for processing, which makes the algorithm fit better into the GPU architecture. We implement our algorithm using the compute unified device architecture (CUDA) on the Nvidia's GTX 650 GPU, which has the latest Kepler architecture. To compare the performance of our algorithm with different data sizes, we built sixteen face AAM models of different dimensional textures. The experiment results show that our parallel AAM fitting algorithm can achieve real-time performance for videos even on very high-dimensional textures.

Facial Feature Detection Algorithm Based on Main Characteristics of Eyes

This paper proposed a new facial feature points localization algorithm based on main characteristics of eyes.Use the result of pupil center position to initialize the model of hybrid improved active shape model (ASM) and active appearance model (AAM). The algorithm will use two-dimensional local gray information to update the feature point position when using ASM to locate the face contour feature points. As to the internal features point location, it establishes facial organs independent AAM model. At the same time, it optimizes measure functions of ASM and AAM to judge the convergence of search algorithm. The experimental results show that the new algorithm greatly improved the localization accuracy of facial feature points.

Active Shape and Active Apparent Hybrid Model of Human Facial Feature Points Localization Algorithm

An improved active shape model (ASM) and active appearance model (AAM) based new method is proposed, this method will use two-dimensional local gray information to update the feature point position when using ASM to locate the face contour feature points. As to the internal features point location, it establishes facial organs independent AAM model. At the same time, it uses different measure functions to judge the convergence of search algorithm. The experimental results show that the new algorithm greatly improved the localization accuracy of facial feature points.

A Two Step Face Alignment Approach Using Statistical Models

Although face alignment using the Active Appearance Model (AAM) is relatively stable, it is known to be sensitive to initial values and not robust under inconstant circumstances. In order to strengthen the ability of AAM performance for face alignment, a two step approach for face alignment combining AAM and Active Shape Model (ASM) is proposed. In the first step, AAM is used to locate the inner landmarks of the face. In the second step, the extended ASM is used to locate the outer landmarks of the face under the constraint of the estimated inner landmarks by AAM. The two kinds of landmarks are then combined together to form the whole facial landmarks. The proposed approach is compared with the basic AAM and the progressive AAM methods. Experimental results show that the proposed approach gives a much more effective performance.

AUTOMATIC SEGMENTATION OF CARDIAC MAGNETIC RESONANCE IMAGES USING ACTIVE APPEARANCE MODELS AND HAUSDORFF DISTANCE

Active Appearance Models (AAM), have been introduced by Cootes et al. [IEEE Transactions on Pattern Analysis and Machine Intelligence, 2001], and are used to learn objects characteristics during a training phase by building a compact statistical model representing shape and texture variation of the object. This Model is used to find the object location and shape-appearance parameters, in a test set. The selection of the initial position of the construct model in a test image is a very important task in this context. The goal of this work is to propose an automatic segmentation method applied to cardiovascular MR images using an AAM based segmentation approach. The AAM model was constructed using 20 end-diastolic and end-systolic short axis cardiac magnetic resonance images (MRI). Once the model is constructed, we select the best position in order to start the search step manually in the test image. That is why; in this paper, the localization of the left ventricular cavity in the test image is used to select the initial position of the construct model developed from the training images. So we propose an automatic approach to detect this spatial position by using two methods: (1) the circular Hough transform (CHT) and (2) the evaluation of the Hausdorff distance.

Performance of active appearance model-based pose-robust face recognition

In this study, the authors study the performance of a novel active appearance model (AAM)-based fully automatic system for pose robust face recognition that allows faster fitting to a no frontal view and generation of virtual views. The system follows a multiresolution scheme, where the first level is used to initialise a generic AAM, pose angle is automatically estimated using eigenvector analysis, and then a pose-dependent AAM model is selected. Next level refines AAM model fitting and registration. Finally, a virtual frontal view is created before face matching. Recognition results over CMU PIE database show similar values compared with the performance achieved with manually landmarked faces. Compared with a classical view-based approach, this multiresolution scheme performs similarly but is sensibly faster.

Smart cameras: 2D affine models for determining subject facial expressions

In this paper we consider how next generation of smart cameras will move beyond face detection and tracking to develop more useful analysis tools for determining the expressions and emotions of the subjects in images. An initial proof-of-concept using an extension of an Active Appearance Model (AAM) to measure facial parameters and a set of classifiers to determine facial states are described. Preliminary findings are encouraging and suggest that the use of the component-based AAM models and SVM classifiers will enable new applications and opportunities in smart cameras.

FACE ALIGNMENT BASED ON STATISTICAL MODELS AND GABOR WAVELETS

This paper presents a novel face alignment method for image sequences that integrates Gabor wavelets with the statistical AAM (Active Appearance Model). First, feature points are characterized using Gabor wavelets and can be individually tracked. A disadvantage of this kind of purely feature-based tracking is that errors accumulate and the feature points loose lock on their corresponding features. To overcome this problem, a face affine transform is used to obtain the initial shape of the AAM model. Finally, the AAM is used to impose global constraints upon the local feature points and to produce an exact alignment. Experimental results demonstrate the ability of the proposed algorithm to accurately align and locate facial features.

Automated Contour Detection in Cardiac MRI Using Active Appearance Models

Definition of the optimal training set for the automated segmentation of short-axis left ventricular magnetic resonance (MR) imaging studies in clinical practice based on active appearance model (AAM). We investigated the segmentation accuracy by varying the size and composition of the training set (ie, the ratio between pathologic and normal ventricle images, and the vendor dependence). The accuracy was assessed using the degree of similarity and the difference in ejection fraction between automatically detected and manually drawn contours. Including more images in the training set results in a better accuracy of the detected contours, with optimum results achieved when including 180 images in the training set. Using AAM-based contour detection with a mixed model of 80% normal-20% pathologic images does provide good segmentation accuracy in clinical routine. Finally, it is essential to define different AAM models for different vendors of MRI systems. A model defined on a sufficient number of images with the correct distribution of image characteristics achieves good matches in clinical routine. It is essential to define different AAM models for different vendors of MRI systems.

Texture enhanced appearance models

Statistical region-based registration methods such as the active appearance model (AAM) are used for establishing dense correspondences in images. At low resolution, images correspondences can be recovered reliably in real-time. However, as resolution increases this becomes infeasible due to excessive storage and computational requirements. We propose to reduce the dimensionality of the textural components by selecting a subset of basis functions from a larger dictionary, estimate regression splines and model only the coefficients of the retained basis functions. We demonstrate the use of two types of bases, namely wavelets and wedgelets. The former extends the previous work of Wolstenholme and Taylor where Haar wavelet coefficient subsets were applied. The latter introduces the wedgelet regression tree based on triangulated domains. The wavelet and wedgelet regression splines are functional descriptions of the intensity information and serve to (1) reduce noise and (2) produce a compact textural description. Dimensionality reduction by subsampling in the CDF 9–7 wavelet and wedgelet representations yield better results than ‘standard’ subsampling in the pixel domain. We show that the bi-orthogonal CDF 9–7 wavelet yields better results than the Haar wavelet. Further, we show that the inherent frequency separation in wavelets allows for cost-free band-pass filtering, e.g. edge-emphasis, and that this edge enhancement provide better results in terms of segmentation accuracy. Wedgelet representation are superior to wavelet representations at high dimensionality-reduction rates. At low reduction rates an edge enhanced wavelet representation provides better segmentation accuracy than the full standard AAM model.