Image Life Trails Based On Contrast Reduction Models For Face Counter-Spoofing

Natural face images are both content and context-rich, in the sense that they carry significant immersive information via depth cues embedded in the form of self-shadows or a space varying blur. Images of planar face prints, on the other hand, tend to have lower contrast and also suppressed depth cues. In this work, a solution is proposed, to detect planar print spoofing by enhancing self-shadow patterns present in face images. This process is facilitated and siphoned via the application of a non-linear iterative functional map, which is used to produce a contrast reductionist image sequence, termed as an image life trail. Subsequent images in this trail tend to have lower contrast in relation to the previous iteration. Differences taken across this image sequence help in bringing out the self-shadows already present in the original image. On a client specific mode, when the subjects and faces are registered, secondary statistics which capture the prominence of self-shadow information, indicate that planar print-images tend to have highly suppressed self-shadows when compared with natural face images. An elaborate tuning procedure, based on a reduced set of training images was developed to first identify the optimal parameter set and then adapt the feature-vectors so that the error-rates were minimized for a specific dataset. Overall mean error rate for the calibration-set (reduced CASIA dataset) was found to be 0.267% and the error rates for other datasets such OULU-NPU and CASIA-SURF were 0.17% and 0.73% respectively

A neural mechanism for detecting object motion during self-motion

Detecting objects that move in a scene is a fundamental computation performed by the visual system. This computation is greatly complicated by observer motion, which causes most objects to move across the retinal image. How the visual system detects scene-relative object motion during self-motion is poorly understood. Human behavioral studies suggest that the visual system may identify local conflicts between motion parallax and binocular disparity cues to depth, and may use these signals to detect moving objects. We describe a novel mechanism for performing this computation based on neurons in macaque area MT with incongruent depth tuning for binocular disparity and motion parallax cues. Neurons with incongruent tuning respond selectively to scene-relative object motion and their responses are predictive of perceptual decisions when animals are trained to detect a moving object during selfmotion. This finding establishes a novel functional role for neurons with incongruent tuning for multiple depth cues.

New Approaches to Visual Scale and Visual Shape

Human 3D vision is thought to triangulate the size, distance, direction, and 3D shape of objects using vision from the two eyes. But all four of these capacities rely on the visual system knowing where the eyes are pointing. Dr Linton's experimental work on size and distance challenge this account, suggesting a purely retinal account of visual size and distance, and likely direction and 3D shape. This requires new accounts of visual scale and visual shape. For visual scale, he argues that observers rely on natural scene statistics to associate accentuated stereo depth (largely from horizontal disparities) with closer distances. This implies that depth / shape is resolved before size and distance. For visual shape, he argues that depth / shape from the two eyes is a solution to a different problem (rivalry eradication between two retinal images treated as if they are from the same viewpoint), rather than the visual system attempting to infer scene geometry (by treating the two retinal images as two different views of the same scene from different viewpoints). Dr Linton also draws upon his book, which questions whether other depth cues (perspective, shading, motion) really have any influence on this process.

Effect of immersive visualization technologies on cognitive load, motivation, usability, and embodiment

Virtual reality (VR) is a promising tool to promote motor (re)learning in healthy users and brain-injured patients. However, in current VR-based motor training, movements of the users performed in a three-dimensional space are usually visualized on computer screens, televisions, or projection systems, which lack depth cues (2D screen), and thus, display information using only monocular depth cues. The reduced depth cues and the visuospatial transformation from the movements performed in a three-dimensional space to their two-dimensional indirect visualization on the 2D screen may add cognitive load, reducing VR usability, especially in users suffering from cognitive impairments. These 2D screens might further reduce the learning outcomes if they limit users’ motivation and embodiment, factors previously associated with better motor performance. The goal of this study was to evaluate the potential benefits of more immersive technologies using head-mounted displays (HMDs). As a first step towards potential clinical implementation, we ran an experiment with 20 healthy participants who simultaneously performed a 3D motor reaching and a cognitive counting task using: (1) (immersive) VR (IVR) HMD, (2) augmented reality (AR) HMD, and (3) computer screen (2D screen). In a previous analysis, we reported improved movement quality when movements were visualized with IVR than with a 2D screen. Here, we present results from the analysis of questionnaires to evaluate whether the visualization technology impacted users’ cognitive load, motivation, technology usability, and embodiment. Reports on cognitive load did not differ across visualization technologies. However, IVR was more motivating and usable than AR and the 2D screen. Both IVR and AR rea ched higher embodiment level than the 2D screen. Our results support our previous finding that IVR HMDs seem to be more suitable than the common 2D screens employed in VR-based therapy when training 3D movements. For AR, it is still unknown whether the absence of benefit over the 2D screen is due to the visualization technology per se or to technical limitations specific to the device.

Nonsymbolic-Magnitude Deficit in Adults With Developmental Dyscalculia: Evidence of Impaired Size Discrimination but Intact Size Constancy

Although researchers have debated whether a core deficit of nonsymbolic representation of magnitude underlies developmental dyscalculia (DD), research has mostly focused on numerosity processing. We probed the possibility of a general magnitude deficit in individuals with DD and asked whether sensitivity to size varied in contexts of depth ordering and size constancy. We measured full psychometric functions in size-discrimination tasks in 12 participants with DD and 13 control participants. Results showed that although people with DD exhibited veridical perceived magnitude, their sensitivity to size was clearly impaired. In contrast, when objects were embedded in depth cues allowing size-constancy computations, participants with DD demonstrated typical sensitivity to size. These results demonstrate a deficit in the perceptual resolutions of magnitude in DD. At the same time, the finding of an intact size constancy suggests that when magnitude perception is facilitated by implicit mandatory computations of size constancy, this deficit is no longer evident.

The Effects of Incorrect Occlusion Cues on the Understanding of Barehanded Referencing in Collaborative Augmented Reality

In many collaborative tasks, the need for joint attention arises when one of the users wants to guide others to a specific location or target in space. If the collaborators are co-located and the target position is in close range, it is almost instinctual for users to refer to the target location by pointing with their bare hands. While such pointing gestures can be efficient and effective in real life, performance will be impacted if the target is in augmented reality (AR), where depth cues like occlusion may be missing if the pointer’s hand is not tracked and modeled in 3D. In this paper, we present a study utilizing head-worn AR displays to examine the effects of incorrect occlusion cues on spatial target identification in a collaborative barehanded referencing task. We found that participants’ performance in AR was reduced compared to a real-world condition, but also that they developed new strategies to cope with the limitations of AR. Our work also identified mixed results of the effect of spatial relationships between users.

Perceptual Integration and Visual Illusions

In my first post I argued that inconsistencies in visual space reflect a conflict between visual experience and perceptual judgement. In this second post I argue that the same approach can be applied to (a) the integration of depth cues, and (b) illusions of visual space, to show that they too operate at the level of cognition rather than perception.

Visual perception and camouflage response to 3-D backgrounds and cast shadows in the European cuttlefish Sepia officinalis

To conceal themselves on the seafloor European cuttlefish Sepia officinalis express a large repertoire of body patterns. Scenes with 3-D relief are especially challenging because neither is it possible to directly recover visual depth from the 2-D retinal image, nor for the cuttlefish to alter its body shape to resemble nearby objects. Here we characterise cuttlefish's camouflage responses to 3-D relief, and to cast shadows, which are complementary depth cues. Animals were recorded in the presence of cylindrical objects of fixed (15mm) diameter, but varying in height, greyscale and strength of cast shadows, and to corresponding 2-D pictorial images. With the cylinders the cuttlefish expressed a ‘3-D’ body pattern, which is distinct from previously described Uniform, Mottle, and Disruptive camouflage patterns. This pattern was insensitive to variation in object height, contrast, and cast shadow, except when shadows were most pronounced, in which case the body patterns resembled those used on the 2-D backgrounds. This suggests that stationary cast shadows are not used as visual depth cues by cuttlefish, and that rather than directly matching the 2-D retinal image, the camouflage response is a two-stage process whereby the animal first classifies the physical environment and then selects an appropriate pattern. Each type of pattern is triggered by specific cues that may compete allowing the animal to select the most suitable camouflage, so the camouflage response is categorical rather than continuously variable. These findings give unique insight into how an invertebrate senses its visual environment to generate the body pattern response.