Computational and Performance Aspects of PCA-Based Face-Recognition Algorithms

Deep learning large-scale drug discovery and repurposing

Large-scale drug discovery and repurposing is challenging. Identifying the mechanism of action (MOA) is crucial, yet current approaches are costly and low-throughput. Here we present an approach for MOA identification by profiling changes in mitochondrial phenotypes. By temporally imaging mitochondrial morphology and membrane potential, we established a pipeline for monitoring time-resolved mitochondrial images, resulting in a dataset comprising 570,096 single-cell images of cells exposed to 1,068 United States Food and Drug Administration-approved drugs. A deep learning model named MitoReID, using a re-identification (ReID) framework and an Inflated 3D ResNet backbone, was developed. It achieved 76.32% Rank-1 and 65.92% mean average precision on the testing set and successfully identified the MOAs for six untrained drugs on the basis of mitochondrial phenotype. Furthermore, MitoReID identified cyclooxygenase-2 inhibition as the MOA of the natural compound epicatechin in tea, which was successfully validated in vitro. Our approach thus provides an automated and cost-effective alternative for target identification that could accelerate large-scale drug discovery and repurposing.

The visual statistical learning overcomes scene dissimilarity through an independent clustering process

Contextual cueing is a phenomenon of visual statistical learning observed in visual search tasks. Previous research has found that the degree of deviation of items from its centroid, known as variability, determines the extent of generalization for that repeated scene. Introducing variability increases dissimilarity between multiple occurrences of the same repeated layout significantly. However, current theories do not explain the mechanisms that help to overcome this dissimilarity during contextual cue learning. We propose that the cognitive system initially abstracts specific scenes into scene layouts through an automatic clustering unrelated to specific repeated scenes, and subsequently uses these abstracted scene layouts for contextual cue learning. Experiment 1 indicates that introducing greater variability in search scenes leads to a hindering in the contextual cue learning. Experiment 2 further establishes that conducting extensive visual searches involving spatial variability in entirely novel scenes facilitates subsequent contextual cue learning involving corresponding scene variability, confirming that learning clustering knowledge precedes the contextual cue learning and is independent of specific repeated scenes. Overall, this study demonstrates the existence of multiple levels of learning in visual statistical learning, where item-level learning can serve as material for layout-level learning, and the generalization reflects the constraining role of item-level knowledge on layout-level knowledge.

A narrow band of image dimensions is critical for face recognition

A key challenge in human and computer face recognition is to differentiate information that is diagnostic for identity from other sources of image variation. Here, we used a combined computational and behavioural approach to reveal critical image dimensions for face recognition. Behavioural data were collected using a sorting and matching task with unfamiliar faces and a recognition task with familiar faces. Principal components analysis was used to reveal the dimensions across which the shape and texture of faces in these tasks varied. We then asked which image dimensions were able to predict behavioural performance across these tasks. We found that the ability to predict behavioural responses in the unfamiliar face tasks increased when the early PCA dimensions (i.e. those accounting for most variance) of shape and texture were removed from the analysis. Image similarity also predicted the output of a computer model of face recognition, but again only when the early image dimensions were removed from the analysis. Finally, we found that recognition of familiar faces increased when the early image dimensions were removed, decreased when intermediate dimensions were removed, but then returned to baseline recognition when only later dimensions were removed. Together, these findings suggest that early image dimensions reflect ambient changes, such as changes in viewpoint or lighting, that do not contribute to face recognition. However, there is a narrow band of image dimensions for shape and texture that are critical for the recognition of identity in humans and computer models of face recognition.

Block-Active ADMM to Minimize NMF with Bregman Divergences

Over the last ten years, there has been a significant interest in employing nonnegative matrix factorization (NMF) to reduce dimensionality to enable a more efficient clustering analysis in machine learning. This technique has been applied in various image processing applications within the fields of computer vision and sensor-based systems. Many algorithms exist to solve the NMF problem. Among these algorithms, the alternating direction method of multipliers (ADMM) and its variants are one of the most popular methods used in practice. In this paper, we propose a block-active ADMM method to minimize the NMF problem with general Bregman divergences. The subproblems in the ADMM are solved iteratively by a block-coordinate-descent-type (BCD-type) method. In particular, each block is chosen directly based on the stationary condition. As a result, we are able to use much fewer auxiliary variables and the proposed algorithm converges faster than the previously proposed algorithms. From the theoretical point of view, the proposed algorithm is proved to converge to a stationary point sublinearly. We also conduct a series of numerical experiments to demonstrate the superiority of the proposed algorithm.

Not the norm: Face likeness is not the same as similarity to familiar face prototypes

Face images depicting the same individual can differ substantially from one another. Ecological variation in pose, expression, lighting, and other sources of appearance variability complicates the recognition and matching of unfamiliar faces, but acquired familiarity leads to the ability to cope with these challenges. Among the many ways that face of the same individual can vary, some images are judged to be better likenesses of familiar individuals than others. Simply put, these images look more like the individual under consideration than others. But what does it mean for an image to be a better likeness than another? Does likeness entail typicality, or is it something distinct from this? We examined the relationship between the likeness of face images and the similarity of those images to average images of target individuals using a set of famous faces selected for reciprocal familiarity/unfamiliarity across US and UK participants. We found that though likeness judgments are correlated with similarity-to-prototype judgments made by both familiar and unfamiliar participants, this correlation was smaller than the correlation between similarity judgments made by different participant groups. This implies that while familiarity weakens the relationship between likeness and similarity-to-prototype judgments, it does not change similarity-to-prototype judgments to the same degree.

Analysing Texture in Portraits

This case study is an initial exploration as to whether the depiction of texture in a set of portraits, all portraying the same Sitter, is related to the familiar likeness assessments reported in a companion paper containing a principal component analysis (PCA) of the portraits' depiction of shape. Somewhat unexpectedly, a texture PCA failed to discriminate the high from low likeness portraits, despite experimentation with different pre-processing methods to reduce the portraits' high level of uninformative, image-level texture variability. There were some findings arising from these analyses, and while only indicative at this stage, include that linear histogram matching is effective in reducing variability in portrait brightness; that depicting, and perhaps exaggerating, shading relating to lighting direction may enhance portrait likeness; and, that whether minimised or exaggerated, lighting direction can be portrayed somewhat anomalously. Furthermore, and in agreement with findings from photographs, shape and texture were not found to be independent variables, and shape-free image registration, while very usefully enabling a comparison of closely corresponding pixel coordinate values, could itself be a confounding factor for undertaking a texture PCA with portraits produced under relatively ambient conditions.

Eco-tourism benefit evaluation of Yellow River based on principal component analysis

In the era of big data transformation, with the emergence of COVID-19, tourism has been given more social responsibilities. Tourism construction in the Yellow River Basin is an indispensable part of tourism construction in China. This paper analyzes the existing eco-tourism resources in Kaifeng City and Shandong Province, as well as the necessity and construction conditions of developing tourism. In this paper, principal component analysis is used to analyze the resource conditions, regional conditions and environmental conditions of the Yellow River tourism resources. The comprehensive evaluation model and index system of tourism resources are constructed. Big data transformation has been realized. The purpose of this paper is to clarify the current situation and potential of tourism in the Yellow River Basin, and to provide reference for the development of tourism in the Yellow River Basin during COVID-19.

Deep Learning of Appearance Affinity for Multi-Object Tracking and Re-Identification: A Comparative View

Recognizing the identity of a query individual in a surveillance sequence is the core of Multi-Object Tracking (MOT) and Re-Identification (Re-Id) algorithms. Both tasks can be addressed by measuring the appearance affinity between people observations with a deep neural model. Nevertheless, the differences in their specifications and, consequently, in the characteristics and constraints of the available training data for each one of these tasks, arise from the necessity of employing different learning approaches to attain each one of them. This article offers a comparative view of the Double-Margin-Contrastive and the Triplet loss function, and analyzes the benefits and drawbacks of applying each one of them to learn an Appearance Affinity model for Tracking and Re-Identification. A batch of experiments have been conducted, and their results support the hypothesis concluded from the presented study: Triplet loss function is more effective than the Contrastive one when an Re-Id model is learnt, and, conversely, in the MOT domain, the Contrastive loss can better discriminate between pairs of images rendering the same person or not.

Cross-View Person Identification Based on Confidence-Weighted Human Pose Matching

Cross-view person identification (CVPI) from multiple temporally synchronized videos taken by multiple wearable cameras from different, varying views is a very challenging but important problem, which has attracted more interest recently. Current state-of-the-art performance of CVPI is achieved by matching appearance and motion features across videos, while the matching of pose features does not work effectively given the high inaccuracy of the 3D pose estimation on videos/images collected in the wild. To address this problem, we first introduce a new metric of confidence to the estimated location of each human-body joint in 3D human pose estimation. Then, a mapping function, which can be hand-crafted or learned directly from the datasets, is proposed to combine the inaccurately estimated human pose and the inferred confidence metric to accomplish CVPI. Specifically, the joints with higher confidence are weighted more in the pose matching for CVPI. Finally, the estimated pose information is integrated into the appearance and motion features to boost the CVPI performance. In the experiments, we evaluate the proposed method on three wearable-camera video datasets and compare the performance against several other existing CVPI methods. The experimental results show the effectiveness of the proposed confidence metric, and the integration of pose, appearance, and motion produces a new state-of-the-art CVPI performance.

Power Quality Disturbance Monitoring and Classification Based on Improved PCA and Convolution Neural Network for Wind-Grid Distribution Systems

The excessive use of power semiconductor devices in a grid utility increases the malfunction of the control system, produces power quality disturbances (PQDs) and reduces the electrical component life. The present work proposes a novel algorithm based on Improved Principal Component Analysis (IPCA) and 1-Dimensional Convolution Neural Network (1-D-CNN) for detection and classification of PQDs. Firstly, IPCA is used to extract the statistical features of PQDs such as Root Mean Square, Skewness, Range, Kurtosis, Crest Factor, Form Factor. IPCA is decomposed into four levels. The principal component (PC) is obtained by IPCA, and it contains a maximum amount of original data as compare to PCA. 1-D-CNN is also used to extract features such as mean, energy, standard deviation, Shannon entropy, and log-energy entropy. The statistical analysis is employed for optimal feature selection. Secondly, these improved features of the PQDs are fed to the 1-D-CNN-based classifier to gain maximum classification accuracy. The proposed IPCA-1-D-CNN is utilized for classification of 12 types of synthetic and simulated single and multiple PQDs. The simulated PQDs are generated from a modified IEEE bus system with wind energy penetration in the balanced distribution system. Finally, the proposed IPCA-1-D-CNN algorithm has been tested with noise (50 dB to 20 dB) and noiseless environment. The obtained results are compared with SVM and other existing techniques. The comparative results show that the proposed method gives significantly higher classification accuracy.

Automated emergency paramedical response system

With the evolution of technology, the fields of medicine and science have also witnessed numerous advancements. In medical emergencies, a few minutes can be the difference between life and death. The obstacles encountered while providing medical assistance can be eliminated by ensuring quicker care and accessible systems. To this effect, the proposed end-to-end system-automated emergency paramedical response system (AEPRS) is semi-autonomous and utilizes aerial distribution by drones, for providing medical supplies on site in cases of paramedical emergencies as well as for patients with a standing history of diseases. Security of confidential medical information is a major area of concern for patients. Confidentiality has been achieved by using decentralised distributed computing to ensure security for the users without involving third-party institutions. AEPRS focuses not only on urban areas but also on semi-urban and rural areas. In urban areas where access to internet is widely available, a healthcare chatbot caters to the individual users and provides a diagnosis based on the symptoms provided by the patients. In semi-urban and rural areas, community hospitals have the option of providing specialised healthcare in spite of the absence of a specialised doctor. Additionally, object recognition and face recognition by using the concept of edge AI enables deep neural networks to run on the edge, without the need for GPU or internet connectivity to connect to the cloud. AEPRS is an airborne emergency medical supply delivery system. It uses the data entered by the user to deduce the best possible solution, in case of an alerted emergency situation and responds to the user accordingly.

Classification of Alzheimer's Disease Stages: An Approach Using PCA-Based Algorithm

Early diagnosis of Alzheimer's disease (AD) allows individuals and their health managers to manage healthier medication. We proposed an approach for classification of AD stages, with respect to principal component analysis (PCA)-based algorithm. The PCA has been extensively applied as the most auspicious face-recognition algorithm. For the proposed algorithm, 100 images of 10 children were transformed for feature extraction and covariance matrix was constructed to obtain eigenvalues. The eigenvector provided a useful framework for face recognition. For the classification of AD stages, magnetic resonance imaging (MRI) and functional magnetic resonance imaging (fMRI) data were obtained from Alzheimer's Disease Neuroimaging Initiative database. Hippocampus is one of the most affected regions by AD. Thus, we selected clusters of voxels from the "hippocampus" of AD screening stage (mild cognitive impairment), AD stage 1, AD stage 2, and AD stage 3. By using eigenvectors corresponding to maximum eigenvalues of fMRI data, the purposed algorithm classified the voxels of AD stages effectively.

Robust Depth-Based Person Re-Identification

Person re-identification (re-id) aims to match people across non-overlapping camera views. So far the RGB-based appearance is widely used in most existing works. However, when people appeared in extreme illumination or changed clothes, the RGB appearance-based re-id methods tended to fail. To overcome this problem, we propose to exploit depth information to provide more invariant body shape and skeleton information regardless of illumination and color change. More specifically, we exploit depth voxel covariance descriptor and further propose a locally rotation invariant depth shape descriptor called Eigen-depth feature to describe pedestrian body shape. We prove that the distance between any two covariance matrices on the Riemannian manifold is equivalent to the Euclidean distance between the corresponding Eigen-depth features. Furthermore, we propose a kernelized implicit feature transfer scheme to estimate Eigen-depth feature implicitly from RGB image when depth information is not available. We find that combining the estimated depth features with RGB-based appearance features can sometimes help to better reduce visual ambiguities of appearance features caused by illumination and similar clothes. The effectiveness of our models was validated on publicly available depth pedestrian datasets as compared to related methods for re-id.

Experimental Demonstration of Feature Extraction and Dimensionality Reduction Using Memristor Networks

Memristors have been considered as a leading candidate for a number of critical applications ranging from nonvolatile memory to non-Von Neumann computing systems. Feature extraction, which aims to transform input data from a high-dimensional space to a space with fewer dimensions, is an important technique widely used in machine learning and pattern recognition applications. Here, we experimentally demonstrate that memristor arrays can be used to perform principal component analysis, one of the most commonly used feature extraction techniques, through online, unsupervised learning. Using Sanger's rule, that is, the generalized Hebbian algorithm, the principal components were obtained as the memristor conductances in the network after training. The network was then used to analyze sensory data from a standard breast cancer screening database with high classification success rate (97.1%).

Learning Invariant Color Features for Person Reidentification

Matching people across multiple camera views known as person reidentification is a challenging problem due to the change in visual appearance caused by varying lighting conditions. The perceived color of the subject appears to be different under different illuminations. Previous works use color as it is or address these challenges by designing color spaces focusing on a specific cue. In this paper, we propose an approach for learning color patterns from pixels sampled from images across two camera views. The intuition behind this work is that, even though varying lighting conditions across views affect the pixel values of the same color, the final representation of a particular color should be stable and invariant to these variations, i.e., they should be encoded with the same values. We model color feature generation as a learning problem by jointly learning a linear transformation and a dictionary to encode pixel values. We also analyze different photometric invariant color spaces as well as popular color constancy algorithm for person reidentification. Using color as the only cue, we compare our approach with all the photometric invariant color spaces and show superior performance over all of them. Combining with other learned low-level and high-level features, we obtain promising results in VIPeR, Person Re-ID 2011, and CAVIAR4REID data sets.

Subject identification via ECG fiducial-based systems: influence of the type of QT interval correction

Electrocardiography (ECG) has been recently proposed as biometric trait for identification purposes. Intra-individual variations of ECG might affect identification performance. These variations are mainly due to Heart Rate Variability (HRV). In particular, HRV causes changes in the QT intervals along the ECG waveforms. This work is aimed at analysing the influence of seven QT interval correction methods (based on population models) on the performance of ECG-fiducial-based identification systems. In addition, we have also considered the influence of training set size, classifier, classifier ensemble as well as the number of consecutive heartbeats in a majority voting scheme. The ECG signals used in this study were collected from thirty-nine subjects within the Physionet open access database. Public domain software was used for fiducial points detection. Results suggested that QT correction is indeed required to improve the performance. However, there is no clear choice among the seven explored approaches for QT correction (identification rate between 0.97 and 0.99). MultiLayer Perceptron and Support Vector Machine seemed to have better generalization capabilities, in terms of classification performance, with respect to Decision Tree-based classifiers. No such strong influence of the training-set size and the number of consecutive heartbeats has been observed on the majority voting scheme.

NUMERICAL ANALYSIS AND COMPARISON OF SPECTRAL DECOMPOSITION METHODS IN BIOMETRIC APPLICATIONS

Face recognition is a challenging problem in computer vision and artificial intelligence. One of the main challenges consists in establishing a low-dimensional feature representation of the images having enough discriminatory power to perform high accuracy classification. Different methods of supervised and unsupervised classification can be found in the literature, but few numerical comparisons among them have been performed on the same computing platform. In this paper, we perform this kind of comparison, revisiting the main spectral decomposition methods for face recognition. We also introduce for the first time, the use of the noncentered PCA and the 2D discrete Chebyshev transform for biometric applications. Faces are represented by their spectral features, that is, their projections onto the different spectral basis. Classification is performed using different norms and/or the cosine defined by the Euclidean scalar product in the space of spectral attributes. Although this constitutes a simple algorithm of unsupervised classification, several important conclusions arise from this analysis: (1) All the spectral methods provide approximately the same accuracy when they are used with the same energy cutoff. This is an important conclusion since many publications try to promote one specific spectral method with respect to other methods. Nevertheless, there exist small variations on the highest median accuracy rates: PCA, 2DPCA and DWT perform better in this case. Also all the covariance-free spectral decomposition techniques based on single images (DCT, DST, DCHT, DWT, DWHT, DHT) are very interesting since they provide high accuracies and are not computationally expensive compared to covariance-based techniques. (2) The use of local spectral features generally provide higher accuracies than global features for the spectral methods which use the whole training database (PCA, NPCA, 2DPCA, Fisher's LDA, ICA). For the methods based on orthogonal transformations of single images, global features calculated using the whole size of the images appear to perform better. (3) The distance criterion generally provides a higher accuracy than the cosine criterion. The use of other p-norms (p > 2) provides similar results to the Euclidean norm, nevertheless some methods perform better. (4) No spectral method can provide 100% accuracy by itself. Therefore, other kind of attributes and supervised learning algorithms are needed. These results are coherent for the ORL and FERET databases. Finally, although this comparison has been performed for the face recognition problem, it could be generalized to other biometric authentication problems.

Evaluation of non-invasive multispectral imaging as a tool for measuring the effect of systemic therapy in Kaposi sarcoma

Diffuse multi-spectral imaging has been evaluated as a potential non-invasive marker of tumor response. Multi-spectral images of Kaposi sarcoma skin lesions were taken over the course of treatment, and blood volume and oxygenation concentration maps were obtained through principal component analysis (PCA) of the data. These images were compared with clinical and pathological responses determined by conventional means. We demonstrate that cutaneous lesions have increased blood volume concentration and that changes in this parameter are a reliable indicator of treatment efficacy, differentiating responders and non-responders. Blood volume decreased by at least 20% in all lesions that responded by clinical criteria and increased in the two lesions that did not respond clinically. Responses as assessed by multi-spectral imaging also generally correlated with overall patient clinical response assessment, were often detectable earlier in the course of therapy, and are less subject to observer variability than conventional clinical assessment. Tissue oxygenation was more variable, with lesions often showing decreased oxygenation in the center surrounded by a zone of increased oxygenation. This technique could potentially be a clinically useful supplement to existing response assessment in KS, providing an early, quantitative, and non-invasive marker of treatment effect.

Color invariants for person reidentification

We revisit the problem of specific object recognition using color distributions. In some applications--such as specific person identification--it is highly likely that the color distributions will be multimodal and hence contain a special structure. Although the color distribution changes under different lighting conditions, some aspects of its structure turn out to be invariants. We refer to this structure as an intradistribution structure, and show that it is invariant under a wide range of imaging conditions while being discriminative enough to be practical. Our signature uses shape context descriptors to represent the intradistribution structure. Assuming the widely used diagonal model, we validate that our signature is invariant under certain illumination changes. Experimentally, we use color information as the only cue to obtain good recognition performance on publicly available databases covering both indoor and outdoor conditions. Combining our approach with the complementary covariance descriptor, we demonstrate results exceeding the state-of-the-art performance on the challenging VIPeR and CAVIAR4REID databases.

Spatial analysis of corresponding fingerprint features from match and close non-match populations

The development of statistical models for forensic fingerprint identification purposes has been the subject of increasing research attention in recent years. This can be partly seen as a response to a number of commentators who claim that the scientific basis for fingerprint identification has not been adequately demonstrated. In addition, key forensic identification bodies such as ENFSI [1] and IAI [2] have recently endorsed and acknowledged the potential benefits of using statistical models as an important tool in support of the fingerprint identification process within the ACE-V framework. In this paper, we introduce a new Likelihood Ratio (LR) model based on Support Vector Machines (SVMs) trained with features discovered via morphometric and spatial analyses of corresponding minutiae configurations for both match and close non-match populations often found in AFIS candidate lists. Computed LR values are derived from a probabilistic framework based on SVMs that discover the intrinsic spatial differences of match and close non-match populations. Lastly, experimentation performed on a set of over 120,000 publicly available fingerprint images (mostly sourced from the National Institute of Standards and Technology (NIST) datasets) and a distortion set of approximately 40,000 images, is presented, illustrating that the proposed LR model is reliably guiding towards the right proposition in the identification assessment of match and close non-match populations. Results further indicate that the proposed model is a promising tool for fingerprint practitioners to use for analysing the spatial consistency of corresponding minutiae configurations.

THE VISUAL-AUDIO INTEGRATED RECOGNITION METHOD FOR USER AUTHENTICATION SYSTEM OF PARTNER ROBOTS

Some of noncontact biometric ways have been used for user authentication system of partner robots, such as visual-based recognition methods and speech recognition. However, the methods of visual-based recognition are sensitive to the light noise and speech recognition systems are perturbed to the acoustic environment and sound noise. Inspiring from the human's capability of compensating visual information (looking) with audio information (hearing), a visual-audio integrating method is proposed to deal with the disturbance of light noise and to improve the recognition accuracy. Combining with the PCA-based and 2DPCA-based face recognition, a two-stage speaker recognition algorithm is used to extract useful personal identity information from speech signals. With the statistic properties of visual background noise, the visual-audio integrating method is performed to draw the final decision. The proposed method is evaluated on a public visual-audio dataset VidTIMIT and a partner robot authentication system. The results verified the visual-audio integrating method can obtain satisfied recognition results with strong robustness.

Feature-specific difference imaging

Difference images quantify changes in the object scene over time. In this paper, we use the feature-specific imaging paradigm to present methods for estimating a sequence of difference images from a sequence of compressive measurements of the object scene. Our goal is twofold. First is to design, where possible, the optimal sensing matrix for taking compressive measurements. In scenarios where such sensing matrices are not tractable, we consider plausible candidate sensing matrices that either use the available a priori information or are nonadaptive. Second, we develop closed-form and iterative techniques for estimating the difference images. We specifically look at l(2)- and l(1)-based methods. We show that l(2)-based techniques can directly estimate the difference image from the measurements without first reconstructing the object scene. This direct estimation exploits the spatial and temporal correlations between the object scene at two consecutive time instants. We further develop a method to estimate a generalized difference image from multiple measurements and use it to estimate the sequence of difference images. For l(1)-based estimation, we consider modified forms of the total-variation method and basis pursuit denoising. We also look at a third method that directly exploits the sparsity of the difference image. We present results to show the efficacy of these techniques and discuss the advantages of each.

Principal component model of multispectral data for near real-time skin chromophore mapping

Multispectral images of skin contain information on the spatial distribution of biological chromophores, such as blood and melanin. From this, parameters such as blood volume and blood oxygenation can be retrieved using reconstruction algorithms. Most such approaches use some form of pixelwise or volumetric reconstruction code. We explore the use of principal component analysis (PCA) of multispectral images to access blood volume and blood oxygenation in near real time. We present data from healthy volunteers under arterial occlusion of the forearm, experiencing ischemia and reactive hyperemia. Using a two-layered analytical skin model, we show reconstruction results of blood volume and oxygenation and compare it to the results obtained from our new spectral analysis based on PCA. We demonstrate that PCA applied to multispectral images gives near equivalent results for skin chromophore mapping and quantification with the advantage of being three orders of magnitude faster than the reconstruction algorithm.

FRVT 2006 and ICE 2006 large-scale experimental results

This paper describes the large-scale experimental results from the Face Recognition Vendor Test (FRVT) 2006 and the Iris Challenge Evaluation (ICE) 2006. The FRVT 2006 looked at recognition from high-resolution still frontal face images and 3D face images, and measured performance for still frontal face images taken under controlled and uncontrolled illumination. The ICE 2006 evaluation reported verification performance for both left and right irises. The images in the ICE 2006 intentionally represent a broader range of quality than the ICE 2006 sensor would normally acquire. This includes images that did not pass the quality control software embedded in the sensor. The FRVT 2006 results from controlled still and 3D images document at least an order-of-magnitude improvement in recognition performance over the FRVT 2002. The FRVT 2006 and the ICE 2006 compared recognition performance from high-resolution still frontal face images, 3D face images, and the single-iris images. On the FRVT 2006 and the ICE 2006 data sets, recognition performance was comparable for high-resolution frontal face, 3D face, and the iris images. In an experiment comparing human and algorithms on matching face identity across changes in illumination on frontal face images, the best performing algorithms were more accurate than humans on unfamiliar faces.