Active Appearance Model Research Articles

Effective Facial Expression Recognition System Based on Hybrid Classification Technique

In our daily life, facial expression recognition has possible functions in various sectors but still it is not understand, because the absence of efficient expression identification methods. Many methods are used to develop the effectiveness of the identification through indicating issues in face detection and extraction aspects in identification expressions. In the first phase, the noise is eliminating from the image by using of preprocessing techniques and to obtain the quality image in order to decrease the computational complexity. The following phase is feature extraction phase. In this phase, we are extracting related features like eyes, mouth and nose. The shape feature of eye part can be extracted by active appearance model (AAM). The texture feature of nose and mouth can be extracted using grey-level co-occurrence matrix (GLCM). Then in the final phase, we have to categorise a facial expression for this categorisation process by introducing an adaptive genetic fuzzy classifier (AGFC) and neural network (NN). Finally, score level fusion of this two classification result will be done to obtain the face emotions.

An Improved Active Contour Method for Medical Image Segmentation using Singular Value Decomposition

An image segmentation method uses dynamic shape models (ACM) executed by methods for level set systems have been successfully used as piece of picture division. essential idea of ACM is to undeniably address shapes zero level game plan of higher dimensional level set limit, & figure progression of frame through improvement of level set.since they always produce sub-regions with continuous boundaries. However, traditional edge-based active contour models have been applicable to only relatively simple images whose sub-regions are uniform without internal edges. A partial solution to the problem of internal edges is to partition an image based on the statistical approach i.e. Information of image intensity measured and active contour model based on level set and Singular Value Decomposition may be applied to improve the efficiency and accuracy in poor quality images. In this paper an Improved Active Contour Method for medical image segmentation is based on active appearance models, active shape models, level set,PCA and Linear Discriminant Analysis are analyzed with intent to produce an inhomogeneous environment using SVD for segmentation of real world images in the presence of intensity in-homogeneity and noise.

Interest emotion recognition approach using self-organising map and motion estimation

Recognising human facial emotions by computer is an interesting problem. Although several approaches have been proposed the recognition rate, amount of used resources and calculation time remain factors for improvement. Our work presents a new approach for recognising basic emotions (joy, sadness, anger, disgust, surprise and fear) in image sequences. We introduced interest emotion and created its corresponding action units (AUs) based on psychological foundations. Our approach is mainly characterised by minimising used data and consequently optimising the computing time and improving the recognition rate. The proposed approach was divided into three steps: face detection using the Viola and Jones method, the extraction of facial features: here we exploited the facial action coding system, which is based on AUs. To detect AUs, we extracted face strategic points using an active appearance model and a block-matching approach. At the last, we classified the results by using the Kohonen self-organising map (SOM).

Detection of the Presence of Safety Helmets on Motorcyclists Using Active Appearance Models

Detection of the Presence of Safety Helmets on Motorcyclists Using Active Appearance Models

Facial expression video analysis for depression detection in Chinese patients

Emotional state analysis of facial expression is an important research content of emotion recognition. At the same time, in the medical field, the auxiliary early screening tools for depression are also urgently needed by clinics. Whether there are differences in facial expression changes between depressive patients and normal people in the same situation, and whether the characteristics can be obtained and recognized from the video images of depressive patients, so as to help doctors detect and diagnose potential depressive patients early are the contents of this study. In this paper, we introduce the videos collection process of depression patients and control group at Shandong Mental Health Center in China. The key facial features are extracted from the collected facial videos by person specific active appearance model. On the basis of locating facial features, we classified depression with the movement changes of eyes, eyebrows and corners of mouth by support vector machine. The results show that these features are effective for automatic classification of depression patients.

Facial Pain Expression Recognition in Real-Time Videos.

Recognition of pain in patients who are incapable of expressing themselves allows for several possibilities of improved diagnosis and treatment. Despite the advancements that have already been made in this field, research is still lacking with respect to the detection of pain in live videos, especially under unfavourable conditions. To address this gap in existing research, the current study proposed a hybrid model that allowed for efficient pain recognition. The hybrid, which consisted of a combination of the Constrained Local Model (CLM), Active Appearance Model (AAM), and Patch-Based Model, was applied in conjunction with image algebra. This contributed to a system that enabled the successful detection of pain from a live stream, even with poor lighting and a low-resolution recording device. The final process and output allowed for memory for storage that was reduced up to 40%–55% and an improved processing time of 20%–25%. The experimental system met with success and was able to detect pain for the 22 analysed videos with an accuracy of 55.75%–100.00%. To increase the fidelity of the proposed technique, the hybrid model was tested on UNBC‐McMaster Shoulder Pain Database as well.

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.

Fully automatic multi-organ segmentation for head and neck cancer radiotherapy using shape representation model constrained fully convolutional neural networks.

Intensity modulated radiation therapy (IMRT) is commonly employed for treating head and neck (H&N) cancer with uniform tumor dose and conformal critical organ sparing. Accurate delineation of organs-at-risk (OARs) on H&N CT images is thus essential to treatment quality. Manual contouring used in current clinical practice is tedious, time-consuming, and can produce inconsistent results. Existing automated segmentation methods are challenged by the substantial inter-patient anatomical variation and low CT soft tissue contrast. To overcome the challenges, we developed a novel automated H&N OARs segmentation method that combines a fully convolutional neural network (FCNN) with a shape representation model (SRM). Based on manually segmented H&N CT, the SRM and FCNN were trained in two steps: (a) SRM learned the latent shape representation of H&N OARs from the training dataset; (b) the pre-trained SRM with fixed parameters were used to constrain the FCNN training. The combined segmentation network was then used to delineate nine OARs including the brainstem, optic chiasm, mandible, optical nerves, parotids, and submandibular glands on unseen H&N CT images. Twenty-two and 10 H&N CT scans provided by the Public Domain Database for Computational Anatomy (PDDCA) were utilized for training and validation, respectively. Dice similarity coefficient (DSC), positive predictive value (PPV), sensitivity (SEN), average surface distance (ASD), and 95% maximum surface distance (95%SD) were calculated to quantitatively evaluate the segmentation accuracy of the proposed method. The proposed method was compared with an active appearance model that won the 2015 MICCAI H&N Segmentation Grand Challenge based on the same dataset, an atlas method and a deep learning method based on different patient datasets. An average DSC = 0.870 (brainstem), DSC = 0.583 (optic chiasm), DSC = 0.937 (mandible), DSC = 0.653 (left optic nerve), DSC = 0.689 (right optic nerve), DSC = 0.835 (left parotid), DSC = 0.832 (right parotid), DSC = 0.755 (left submandibular), and DSC = 0.813 (right submandibular) were achieved. The segmentation results are consistently superior to the results of atlas and statistical shape based methods as well as a patch-wise convolutional neural network method. Once the networks are trained off-line, the average time to segment all 9 OARs for an unseen CT scan is 9.5 s. Experiments on clinical datasets of H&N patients demonstrated the effectiveness of the proposed deep neural network segmentation method for multi-organ segmentation on volumetric CT scans. The accuracy and robustness of the segmentation were further increased by incorporating shape priors using SMR. The proposed method showed competitive performance and took shorter time to segment multiple organs in comparison to state of the art methods.

Deep Appearance Models: A Deep Boltzmann Machine Approach for Face Modeling

The "interpretation through synthesis" approach to analyze face images, particularly Active Appearance Models (AAMs) method, has become one of the most successful face modeling approaches over the last two decades. AAM models have ability to represent face images through synthesis using a controllable parameterized Principal Component Analysis (PCA) model. However, the accuracy and robustness of the synthesized faces of AAM are highly depended on the training sets and inherently on the generalizability of PCA subspaces. This paper presents a novel Deep Appearance Models (DAMs) approach, an efficient replacement for AAMs, to accurately capture both shape and texture of face images under large variations. In this approach, three crucial components represented in hierarchical layers are modeled using the Deep Boltzmann Machines (DBM) to robustly capture the variations of facial shapes and appearances. DAMs are therefore superior to AAMs in inferencing a representation for new face images under various challenging conditions. The proposed approach is evaluated in various applications to demonstrate its robustness and capabilities, i.e. facial super-resolution reconstruction, facial off-angle reconstruction or face frontalization, facial occlusion removal and age estimation using challenging face databases, i.e. Labeled Face Parts in the Wild (LFPW), Helen and FG-NET. Comparing to AAMs and other deep learning based approaches, the proposed DAMs achieve competitive results in those applications, thus this showed their advantages in handling occlusions, facial representation, and reconstruction.

A System Automation using Human Eye Motion Based on Active Appearance Model(AAM)

International Journal of Computer Sciences and Engineering (A UGC Approved and indexed with DOI, ICI and Approved, DPI Digital Library) is one of the leading and growing open access, peer-reviewed, monthly, and scientific research journal for scientists, engineers, research scholars, and academicians, which gains a foothold in Asia and opens to the world, aims to publish original, theoretical and practical advances in Computer Science,Information Technology, Engineering (Software, Mechanical, Civil, Electronics & Electrical), and all interdisciplinary streams of Computing Sciences. It intends to disseminate original, scientific, theoretical or applied research in the field of Computer Sciences and allied fields. It provides a platform for publishing results and research with a strong empirical component. It aims to bridge the significant gap between research and practice by promoting the publication of original, novel, industry-relevant research.

Deep appearance models for face rendering

We introduce a deep appearance model for rendering the human face. Inspired by Active Appearance Models, we develop a data-driven rendering pipeline that learns a joint representation of facial geometry and appearance from a multiview capture setup. Vertex positions and view-specific textures are modeled using a deep variational autoencoder that captures complex nonlinear effects while producing a smooth and compact latent representation. View-specific texture enables the modeling of view-dependent effects such as specularity. In addition, it can also correct for imperfect geometry stemming from biased or low resolution estimates. This is a significant departure from the traditional graphics pipeline, which requires highly accurate geometry as well as all elements of the shading model to achieve realism through physically-inspired light transport. Acquiring such a high level of accuracy is difficult in practice, especially for complex and intricate parts of the face, such as eyelashes and the oral cavity. These are handled naturally by our approach, which does not rely on precise estimates of geometry. Instead, the shading model accommodates deficiencies in geometry though the flexibility afforded by the neural network employed. At inference time, we condition the decoding network on the viewpoint of the camera in order to generate the appropriate texture for rendering. The resulting system can be implemented simply using existing rendering engines through dynamic textures with flat lighting. This representation, together with a novel unsupervised technique for mapping images to facial states, results in a system that is naturally suited to real-time interactive settings such as Virtual Reality (VR).

Face aging simulation with a new wrinkle oriented active appearance model

The use of computer simulation to understand how human faces age has been a growing area of research since decades. It has been applied to the search for missing children as well as to the fields of entertainment, cosmetics and dermatology research. Our objective is to elaborate a model for the age-related changes of visual cues which affect the perception of age, so that we may better predict them. Traditional approaches based on the Active Appearance Model (AAM) tend to blurry appearance and wipe out texture details such as wrinkles. We introduce Wrinkle Oriented Active Appearance Model (WOAAM) where a new channel is added to the AAM dedicated to analyze wrinkles. Firstly, we propose to represent both the shape and texture of each wrinkle on a face by a compact and interpretable vector. Afterwards, to model the distribution of wrinkles on a face, we introduce a new way to approximate an empiric joint probability density by creating an ensemble of joint probability densities estimated by Kernel Density Estimation. Finally, we show how to create new samples from such an ensemble of densities, and thus synthesize new plausible wrinkles. In comparison to other methods which add wrinkles at post-processing level, our method fully integrates them in AAM. Thereby, the wrinkles generated are statistically representative of a specific age in terms of number, length, shape and intensity. With an age estimation Convolutional Neural Network, we found that age-progressed faces produced by the WOAAM better reduces the gap between the expected age and the estimated age than those produced by a classic AAM.

Facial age feature extraction based on deep sparse representation

In recent years, the research on facial age estimation has attracted more and more interests from the scholars and researchers in the field of psychological, aesthetic, forensic science and computer vision. However, facial age estimation is a challenging work: face age does not suffer from only the influence of intrinsic factors (e.g., genes) but also external factors (e.g., living conditions), so that it is difficult to find accurate features which can describe the change of age. Therefore, this paper proposes a robust face feature extraction method based on dynamic deep sparse representation. Combined with the respective characteristics of Active Appearance Model (AAM), Local Binary Patterns (LBP), Gabor and Bio-Inspired Features (BIF), this method will give sufficient consideration to the thinking way of object recognition of the human, similarity of adjacent ages and the classification principle of signal sparse representation. In addition, in order to reduce the interference of the face identity factor, two factors analysis method is proposed to separate the face identity factor. The experimental results show that the feature extraction method proposed in this paper has strong discrimination and robustness, which outperforms the state-of-the-art age estimation approaches.

Automatic segmentation of puborectalis muscle on three-dimensional transperineal ultrasound.

ABSTRACTObjectivesThe introduction of three‐dimensional (3D) analysis of the puborectalis muscle (PRM) for diagnostic purposes into daily practice is hindered by the need for appropriate training of observers. Automatic segmentation of the PRM on 3D transperineal ultrasound may aid its integration into clinical practice. The aims of this study were to present and assess a protocol for manual 3D segmentation of the PRM on 3D transperineal ultrasound, and to use this for training of automatic 3D segmentation method of the PRM.MethodsThe data used in this study were derived from 3D transperineal ultrasound sequences of the pelvic floor acquired at 12 weeks' gestation from nulliparous women with a singleton pregnancy. A manual 3D segmentation protocol was developed for the PRM based on a validated two‐dimensional segmentation protocol. For automatic segmentation, active appearance models of the PRM were developed, trained using manual segmentation data from 50 women. The performances of both manual and automatic segmentation were analyzed by measuring the overlap and distance between the segmentations. Intraclass correlation coefficients (ICCs) and their 95% CIs were determined for mean echogenicity and volume of the puborectalis muscle, in order to assess inter‐ and intraobserver reliabilities of the manual method using data from 20 women, as well as to compare the manual and automatic methods.ResultsInterobserver reliabilities for mean echogenicity and volume were very good for manual segmentation (ICCs 0.987 and 0.910, respectively), as were intraobserver reliabilities (ICCs 0.991 and 0.877, respectively). ICCs for mean echogenicity and volume were very good and good, respectively, for the comparison of manual vs automatic segmentation (0.968 and 0.626, respectively). The overlap and distance results for manual segmentation were as expected, showing an average mismatch of only 2–3 pixels and reasonable overlap. Based on overlap and distance, five mismatches were detected for automatic segmentation, resulting in an automatic segmentation success rate of 90%.ConclusionsThis study presents a reliable manual segmentation protocol and automatic 3D segmentation method for the PRM, which will facilitate future investigation of the PRM, allowing for the reliable measurement of potentially clinically valuable parameters such as mean echogenicity. © 2017 The Authors. Ultrasound in Obstetrics & Gynecology Published by John Wiley & Sons Ltd on behalf of the International Society of Ultrasound in Obstetrics and Gynecology.

Content and context based image retrieval classification based on firefly-neural network

In this study presents an improved classification and feature reduction techniques are newly introduced in our proposed search area of image retrieval. Especially we are including a shape feature that has been so active and successful in the past few years. In our proposed approach the texture feature is extracted by using improved multi texton technique and GLCM technique and the textual features are keywords, annotations. The Visual features are extracted with the aid of a bag of visual words (BOW) also it can be approved by employ Scaling Invariant Feature Transform (SIFT) and Fuzzy C-Means Clustering technique the features are edge detection, corner detection. Improving a performance and accuracy newly we are including a shape based features the features are Active Appearance Model (AAM) and shape surface plot (SURF). In the second phase, we are including feature reduction method is applied to reduce the features’ space without losing the accuracy of classification in this reduction phases we are including an OLPP technique to perform a feature reduction phases. After that, the images are classified with the aid of modified neural network algorithm. This will classify the images in the different class in order to improve the precision and recall rate. After classification, we are including a squared Euclidian distance method to retrieve minimum distance image the similar images are retrieved from the relevant class as per the given query image.

Robust facial expression recognition using RGB-D images and multichannel features

Traditional and classical methods of facial expression recognition are mainly based on intensity image and are prone to be disturbed by illumination, poses, and disguise. This research aims to develop a robust facial expression recognition method using RGB-D images and multichannel features. Based on image entropy and visual saliency, facial texture features are firstly extracted from RGB images and depth images to construct the Histogram of Oriented Gradient (HOG) descriptors. And then, we extract geometric features of RGB images using Active Appearance Model (AAM). Combining the HOG texture features with the AAM geometric feature, we build a robust multichannel feature vector for facial expression recognition. On this basis, an improved Support Vector Machine (SVM) algorithm, namely GS-SVM, is used to classify facial expression recognition. The proposed GS-SVM algorithm applies Grid Search method to optimize the best parameters for SVM classifier and estimate the accuracy of each parameter combination in specified range. Finally, the proposed methods are tested and evaluated on the merged RGB-D database. Experimental results show that the proposed algorithm not only achieves a higher average recognition rate but also is robust to uncontrolled environments.

The relationship between accurate 3D measurements of cartilage thickness and radiographic joint space width: data from the osteoarthritis initiative

Purpose: Radiographic joint space width (rJSW) measurement has historically been the preferred measure of disease progression in OA structure modification trials. Magnetic resonance imaging (MRI) can visualise change in multiple tissues, and offers increased sensitivity to change, allowing for shorter trials with fewer patients. In order to compare changes measured using MRI biomarkers with previous studies using rJSW, it is important to understand (1) how well the new biomarkers agree with radiographic measures, and (2) the contribution made by different knee tissues to changes in rJSW. This is challenging, as MR images are normally obtained with the patient in a supine position, while radiographs are taken in a standing position, with the knee slightly flexed. We aimed to assess agreement between accurate 3D MRI measurements of cartilage thickness in the medial tibiofemoral joint and quantified rJSW, in a large dataset. Methods: Analyses were performed on all knees from the Osteoarthritis Initiative at baseline, which had a KL grade and a recorded quantitative measure of joint space width (5726 knees). Cartilage in the central medial regions of the femur and tibia were automatically segmented using a method based on active appearance models, a form of statistical shape model. This method produces highly accurate segmentations (mean accuracy = 0.05 mm, 95th percentile = 0.45, meaning that all errors are less than the average edge length of one voxel). Average thickness of the central medial femur and tibia regions were added to provide a measure of total cartilage thickness. Radiographic JSW measures were provided by the OAI, and were generated using the algorithm-based method of Duryea; for this comparison, we selected a position approximately in the middle of the tibial plateau, also found to be the most responsive in clinical studies (JSW 225). Agreement between the methods was assessed using Bland Altman plots. Results: For KL grades 0, 1 and 2, the systematic bias between rJSW and cartilage thickness was around 1mm, the 95% confidence limits were 1, 1.2 and 1.26 mm respectively (Figure 1). Figure 1 also shows typical examples of MR images for each KL grade within the region in which MR and radiographic measures are taken. For KL grades 0–2, the agreement between measures is close to theoretical agreement limits, which we estimate to be 0.74 mm based on the measurement errors for the methods. In KL3 knees, the bias reduces to 0.4 mm, and the 95% CL increases to 1.69. In KL4 knees, the bias is almost eliminated (0.14 mm). Conclusions: rJSW is composed of various pathologies that are not visualised in radiographs, which consolidate all changes into a single projection. These include cartilage loss, meniscal extrusion and synovial fluid. In healthy knees, and in the early stages of the disease (KL 2), agreement between cartilage thickness and rJSW is excellent – as close as realistically possible in measures of biological systems. The systematic bias of 1mm in KL grade 0–2 knees is most likely contributed by the posterior portion of the medial meniscus which is located within the field of measurement. The decrease in bias at KL3 is probably caused by the increasing damage seen in the menisci, which are less able to keep the bones separate when the patient is standing. At KL4, the meniscus is probably incapable of keeping the bones separate when loaded. The high value for the 95% confidence limits at KL4 are likely due to multiple factors within the knee, including meniscal damage and changes in bone shape, together with the inability of the rJSW method to discriminate the bone edges, once they are very close to each other. Additionally, it can be difficult to position the flexed knee in the radiograph both accurately and reproducibly when there is no clear joint space width, and bone shapes within the joint are altered. There may in fact be 2 groups within the KL4 Bland Altman plot - one group towards the right probably still maintains some joint space width, and perhaps a more competent meniscus, and retains the bias of KL0 to KL3 knees.

The relationship between two different measures of osteoarthritis bone pathology, bone marrow lesions and 3D bone shape: data from the osteoarthritis initiative

Purpose: Magnetic resonance imaging (MRI) has provided great insights into the development of OA, and helped highlight the importance of subchondral bone pathology. Bone shape and bone marrow lesions (BMLs) represent different features of MRI-detected subchondral pathology in osteoarthritis (OA) and while BMLs are well studied the relationship between these two remains poorly understood. It is important to understand if they represent a single construct or different parts of the OA process, and to further explore their use as imaging biomarkers. The aim of this study was to determine how these features are related and how they change in OA progression. Methods: 600 participants from the FNIH Biomarkers Initiative of the OAI were included, having Kellgren–Lawrence (KL) 1, 2 or 3 at baseline and MRI data at baseline and 24 months. MRI images obtained from the OAI were scored for BMLs using the semi-quantitative MOAKS system that scores 3 features using an ordinal scale for size, number of BMLs and percentage of lesion that is a BML as opposed to a cyst. Quantitative 3D bone shape data was provided by Imorphics (Manchester, UK) using active appearance models (AAMs) applied separately for femur and tibia from 3T DESS-we images. The shape vector for each bone was calculated by taking the principal components of the mean non-OA shape and the mean OA shape and the vector was normalised by setting the mean OA shape to +1 and mean non-OA shape as −1, with each unit increase of 1 representing 1 standard deviation (SD) calculated in non-OA knees. Four BML scores were computed: “BML total size”, “BML total number”, “BML total sub regions” and BML maximum size”. At baseline the correlation between the BML scores and bone shape measures were assessed using Spearman`s correlation coefficients. The association between 3D quantitative bone shape vectors and BMLs at baseline was analysed using linear regression. Presence of a BML was defined as BML total size ≥ 1. Responsiveness was calculated for both features in 4 pre-defined progression groups and reported as standardised response means (SRMs).Due to nesting of repeated measurements within individuals, multilevel linear modelling was used to assess the longitudinal relationship between change in BML size and change in bone shape vector. Results: Mean age was 61.5, 59% were female and mean BMI was 30.7. Correlation between baseline femur shape vector and femur BML total size was r = 0.28, P < 0.001 and similar for the tibia. Linear models revealed an association between presence of a femur BML at baseline and baseline 3D femur vector in both univariable and multivariable models (adjusted coefficient 0.49, 95% CI 0.30, 0.68) indicating a more positive femur vector (indicative of “increased OA”) in individuals with BMLs at baseline, with a difference equivalent to 0.5 × SD of non-OA knees. Longitudinally bone shape demonstrated more responsiveness to change than all BML measures (SRM 0.89 vs 0.13) for femur bone shape vector vs BML total size respectively (Table 1). Univariable growth models revealed that bone shape vector tended to be more positive over time (indicating worsening) while BML total size reduced over time. Multilevel models also showed that an increase in BML size was related to a more positive bone shape vector (representing worsening OA). Conclusions: There is a relationship between bone shape and BMLs, with increasing OA bone shape vector associated with higher prevalence of BMLs. Bone shape vector demonstrated greater responsiveness than semi-quantitative BMLs over time periods typical of a clinical trial.Table 1Responsiveness of bone shape and BML measures over 24 monthsRadiographic and pain progression N = 194Radiographic progression only N = 103Pain progression only N = 103No progression N = 2002 year responsiveness, SRM (95%CI)Bone shapeFemur0.89 (0.72,1.02)1.02 (0.85,1.20)0.46 (0.31,0.61)0.61 (0.49,0.72)Tibia0.84 (0.70,0.97)0.76 (0.56,0.96)0.26 (0.07,0.43)0.47 (0.33,0.69)BMLsFemur BML total size−0.13 (−0.26,0.02)−0.15 (−0.35,0.07)−0.31 (−0.51,−0.13)−0.24 (−0.37,0.13)Femur BML total number0.38 (0.26,0.50)0.20 (−0.02,0.41)0.17 (−0.04,0.35)0.27 (0.14,0.38)Tibia BML total size0.11 (−0.02,0.26)0.14 (−0.04,0.31)−0.04 (−0.23,0.15)−0.01(−0.16,0.11)Tibia BML total number0.37 (0.24,0.51)0.31 (0.12,0.51)0.19 (0.00,0.33)0.16 (0.02,0.29)Femur BML maximum size−0.05 (−0.19,−0.09)−0.01 (−0.19,0.21)−0.20 (−0.38,0.01)−0.11 (−0.25,0.02)BML total sub regions−0.02 (−0.15,0.13)−0.03 (−0.23,0.15)−0.13 (−0.32,0.07)−0.06 (−0.20,0.08) Open table in a new tab

Feature extraction for face recognition via Active Shape Model (ASM) and Active Appearance Model (AAM)

Biometric is a pattern recognition system which is used for automatic recognition of persons based on characteristics and features of an individual. Face recognition with high recognition rate is still a challenging task and usually accomplished in three phases consisting of face detection, feature extraction, and expression classification. Precise and strong location of trait point is a complicated and difficult issue in face recognition. Cootes proposed a Multi Resolution Active Shape Models (ASM) algorithm, which could extract specified shape accurately and efficiently. Furthermore, as the improvement of ASM, Active Appearance Models algorithm (AAM) is proposed to extracts both shape and texture of specified object simultaneously. In this paper we give more details about the two algorithms and give the results of experiments, testing their performance on one dataset of faces. We found that the ASM is faster and gains more accurate trait point location than the AAM, but the AAM gains a better match to the texture.

AAM Based Face Sketch Synthesis

Face sketch synthesis has many practical applications, such as law enforcement and digital entertainment. Existing face sketch synthesis methods focus on neighbor selection and/or weight reconstruction. However, these approaches did not take “interpretation through synthesis” into consideration obviously. Active appearance model (AAM) is one of “interpretation through synthesis” approaches. In this paper, we introduce AAM to “explain” face photos by generating synthetic images that are as similar as possible. Then AAM provides a compact set of parameters that are useful for face sketch synthesis. Extensive experiments on public face sketch databases demonstrate the superiority of the proposed method in comparison to state-of-the-art methods.