Face Detection Task Research Articles

Face detection in normal and prosopagnosic individuals

Face detection, the process of finding a face in a visual scene, is a critical step in face processing, yet it has received relatively little attention compared with other face processes. The present study addresses this crucial first stage by investigating the effect of inversion on face detection and by examining how individuals with developmental prosopagnosia perform on face detection tasks. Fourteen control participants and fourteen individuals with developmental prosopagnosia (DPs) were tested with two face detection tasks: (I) Face versus Non-Face, where arrays of small images were presented, one of which could contain a face and (2) Face versus Face Parts, where a two-tone face could be embedded in a larger array of similar two-tone face parts. On each trial, participants made a speeded response if a face was present in the visual display. On almost all measures both normal and prosopagnosic individuals showed strong inversion effects with significantly worse performance with inverted faces. This shows that the simple task of detection can show inversion effects comparable to those seen for other face tasks, including recognition. Finally, while there were prosopagnosics who were well within the normal range for detection, there were significant group differences, particularly for the case of the Face versus Face Parts, where prosopagnosics were worse than controls on upright but not on inverted face detection.

480-GMACS/mW Resonant Adiabatic Mixed-Signal Processor Array for Charge-Based Pattern Recognition

A resonant adiabatic mixed-signal VLSI array delivers 480 GMACS (10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">9</sup> multiply-and-accumulates per second) throughput for every mW of power, a 25-fold improvement over the energy efficiency obtained when resonant clock generator and line drivers are replaced with static CMOS drivers. Losses in resonant clock generation are minimized by activating switches between the LC tank and DC supply with a periodic pulse signal, and by minimizing the variability of the capacitive load to maintain resonance. We show that minimum energy is attained for relatively wide pulse width, and that typical load distribution in template-based charge-mode computation implies almost constant capacitive load. The resonantly driven 256 times 512 array of 3-T charge-conserving multiply-accumulate cells is embedded in a template matching processor for image classification and validated in a face detection task.

Accuracy versus speed in context-based object detection

The visual detection and recognition of objects is facilitated by context. This paper studies two types of learning methods for realizing context-based object detection in paintings. The first method is called the gradient method; it learns to transform the spatial context into a gradient towards the object. The second method, the context-detection method, learns to detect image regions that are likely to contain objects. The accuracy and speed of both methods are evaluated on a face-detection task involving natural and painted faces in a wide variety of contexts. The experimental results show that the gradient method enhances accuracy at the cost of computational speed, whereas the context-detection method optimises speed at the cost of accuracy. The different results of both methods are argued to arise from the different ways in which the methods trade-off accuracy and speed. We conclude that both the gradient method and the context-detection method can be applied to reliable and fast object detection in paintings and that the choice for either method depends on the application and user constraints.

Early visual components (P100, N170) are disrupted in chronic schizophrenic patients: an event-related potentials study

Introduction On the basis of the Positive and Negative Syndrome Scale (PANSS), fourteen schizophrenic patients and 7 normal controls were confronted with pictures from the Ekman & Friesen series in an event-related potentials study. Procedure Participants were confronted with a visual face-detection task, in which they had to detect, as quickly as possible, deviant faces amongst a train of standard stimuli (neutral faces). Deviant faces changed either on identity (different identity, neutral expression), or on emotion (same identity, happy, fearful or sad expression). Results Schizophrenics exhibited a decrease in amplitude of the face N170, recorded around 170 ms at occipito-temporal sites; this was observed as well for emotional as for identity faces, which suggests a global involvement of face processing. Moreover, this decrease of the face-N170 was positively correlated to positive, but not negative, symptoms of schizophrenia. Finally, the amplitude of P100 was also decreased, which suggests that the N170 decrement would result from a more global deficit in visual processing deficit. Discussion It is suggested that, in schizophrenics, an involvement of early visual processing might underlie the decreased amplitudes and the higher onset latencies of later P300 and N400 components.

A robust automatic face recognition system for real‐time personal identification

PurposeAims to develop an efficient and robust system for real‐time personal identification by automatic face recognition.Design/methodology/approachA wavelet‐based image hierarchy and a guided coarse‐to‐fine search scheme are introduced to improve the computation efficiency in the face detection task. In addition, a Gabor‐based low feature dimensional pattern is proposed to deal with the face recognition problem.FindingsThe proposal of a wavelet‐based image hierarchy and a guided coarse‐to‐fine search scheme is effective to improve the computation efficiency in the face detection task. The introduction of a low feature dimensional pattern is powerful to cope with the transformed appearance‐based face recognition problem. In addition, the use of aggregated Gabor filter responses to represent face images provides a better solution to face feature extraction.Research limitations/implicationsProvides guidance in the design of automatic face recognition system for real‐time personal identification.Practical implicationsBiometrics recognition has been emerging as a new and effective identification technology that attains certain level of maturity. Among many body characteristics that have been used, face is one of the most commonly used characteristics and has drawn considerably large attentions. An automated system to confirm an individual's identity employing features of face is very attractive in many specialized fields.Originality/valueIntroduces a wavelet‐based image hierarchy and a guided coarse‐to‐fine search scheme to improve the computation efficiency in the face detection task. Introduces a Gabor‐based low feature dimensional pattern to deal with the face recognition problem.

Cue-Guided Search: A Computational Model of Selective Attention

Selective visual attention in a natural environment can be seen as the interaction between the external visual stimulus and task specific knowledge of the required behavior. This interaction between the bottom-up stimulus and the top-down, task-related knowledge is crucial for what is selected in the space and time within the scene. In this paper, we propose a computational model for selective attention for a visual search task. We go beyond simple saliency-based attention models to model selective attention guided by top-down visual cues, which are dynamically integrated with the bottom-up information. In this way, selection of a location is accomplished by interaction between bottom-up and top-down information. First, the general structure of our model is briefly introduced and followed by a description of the top-down processing of task-relevant cues. This is then followed by a description of the processing of the external images to give three feature maps that are combined to give an overall bottom-up map. Second, the development of the formalism for our novel interactive spiking neural network (ISNN) is given, with the interactive activation rule that calculates the integration map. The learning rule for both bottom-up and top-down weight parameters are given, together with some further analysis of the properties of the resulting ISNN. Third, the model is applied to a face detection task to search for the location of a specific face that is cued. The results show that the trajectories of attention are dramatically changed by interaction of information and variations of cues, giving an appropriate, task-relevant search pattern. Finally, we discuss ways in which these results can be seen as compatible with existing psychological evidence.

Neural network face detection

Analysing a video sequence is a challenge, because faces are constantly in dynamic motion, presenting many different possible rotational and illumination conditions. While solutions to the task of face detection have been presented, the detection performances of many systems are heavily dependent upon a strictly constrained environment. This paper presents the results of an image-based neural network face detection system which seeks to address the problem of detecting faces under gross variations.

The many faces of the gamma band response to complex visual stimuli

While much is known about the functional architecture of the visual system, little is known about its large-scale dynamics during perception. This study describes this dynamics with a high spatial, temporal and spectral resolution. We recorded depth EEG of epileptic patients performing a face detection task and found that the stimuli induced strong modulations in the gamma band (40 Hz to 200 Hz) in selective occipital, parietal and temporal sites, in particular the fusiform gyrus, the lateral occipital gyrus and the intra-parietal sulcus. Occipito-temporal sites were the first to be activated, closely followed by the parietal sites, while portions of the primary visual cortex seemed to deactivate temporarily. Some of those effects were found to be correlated across distant sites, suggesting that a coordinated balance between regional gamma activations and deactivations could be involved during visual perception.

The Use of Neural Networks in Real-time Face Detection

As continual research is being conducted in the area of computer vision, one of the most practical applications under vigorous development is in the construction of a robust real-time face detection system. Successfully constructing a real-time face detection system not only implies a system capable of analyzing video streams, but also naturally leads onto the solution to the problems of extremely constraint testing environments. Analyzing a video sequence is the current challenge since faces are constantly in dynamic motion, presenting many different possible rotational and illumination conditions. While solutions to the task of face detection have been presented, detection performances of many systems are heavily dependent upon a strictly constrained environment. The problem of detecting faces under gross variations remains largely uncovered. This study presents a real-time face detection system which uses an image based neural network to detect images.

Efficient face candidates selector for face detection

In this paper an efficient face candidates selector is proposed for face detection tasks in still gray level images. The proposed method acts as a selective attentional mechanism. Eye-analogue segments at a given scale are discovered by finding regions which are roughly as large as real eyes and are darker than their neighborhoods. Then a pair of eye-analogue segments are hypothesized to be eyes in a face and combined into a face candidate if their placement is consistent with the anthropological characteristic of human eyes. The proposed method is robust in that it can deal with illumination changes and moderate rotations. A subset of the FERET data set and the BioID face database are used to evaluate the proposed method. The proposed face candidates selector is successful in 98.75% and 98.6% cases, respectively.

Configural face processes in acquired and developmental prosopagnosia: evidence for two separate face systems?

Configural face processes were tested using face recognition and face detection tasks in a comparison of acquired and developmental prosopagnosia. In the recognition task the two patients showed a very different pattern. The developmental patient does not show an inversion effect while the acquired prosopagnosia patient is better at matching inverted than normal stimuli. Moreover, there is no effect of face context on matching features in the developmental case while the acquired prosopagnosia patient shows a strong negative effect of context. However, in a speeded face detection task both patients are similarly unimpaired. The results are consistent with the existence of two separate face systems, one involved in face detection and the other in face recognition.

Chromatic Cues for Face Detection in Natural Scenes

One of the challenging problems of human and machine vision is the detection of objects against complex backgrounds. Our research addresses the question of how faces can be very quickly detected in naturalistic scenes on the basis of luminance and chromatic cues. Although luminance information varies with pose and illumination differences, chromatic information is by and large invariant under these transformations. Hence, chromatic information might be a very powerful cue for segmentation and detection. We compared faces of different pigmentation against background scenes of different colours. Specifically, colour histograms were computed in a perceptually uniform colour space (L*u*v*). We computed the Euclidian distances between the averages of the colour histograms of faces and scenes in L*u*v*. This metric was used to calibrate the contrast between face and scene colour in the experimental design. In a face detection task, subjects saw faces against scene backgrounds at a different distance in colour space. Each combination face - scene was presented for 120 ms (to prevent saccadic explorations), and the subject's task was to indicate whether or not a face was present. Controls involved face - scene pairs on an isoluminant background. Results revealed that luminance information did not affect detection on the basis of chromatic cues. Importantly, the metric of detectability in L*u*v* space between scene and faces predicted reaction times to detection.

Face Detection in Natural Scenes

The first task of any face processing system is detection of the face. We studied how the human visual system achieves face detection using a 2AFC task in which subjects were required to detect a face in the image of a natural scene. Luminance noise was added to the stimuli and performance was measured as a function of orientation and orientation bandwidth of the noise. Sensitivity levels and the effects of orientation bandwidth were similar for horizontally and vertically oriented noise. Performance was reduced for the smallest orientation bandwidth (5.6°) noise but sensitivity did not decrease further with increasing bandwidth until a point between 45° and 90°. The results suggest that important information may be oriented close to the vertical and horizontal. To test whether the results were specific to the task of face detection the same noise was added to the images in a man-made natural decision task. Equivalent levels of noise were found to be more disruptive and the effect of orientation bandwidth was different. The results are discussed in terms of models of face processing making use of oriented filters (eg Watt and Dakin, 1993 Perception22 Supplement, 12) and local energy models of feature detection (Morrone and Burr, 1988 Proceedings of the Royal Society of London B235 221 – 245).

HUMAN FACE DETECTION USING SILHOUETTES

Face detection is integral to any automatic face recognition system. The goal of this research is to develop a system that performs the task of human face detection automatically in a scene. A system to correctly locate and identify human faces will find several applications, some examples are criminal identification and authentication in secure systems. This work presents a new approach based on principal component analysis. Face silhouettes instead of intensity images are used for this research. It results in reduction in both space and processing time. A set of basis face silhouettes are obtained using principal component analysis. These are then used with a Hough-like technique to detect faces. The results show that the approach is robust, accurate and reasonably fast.