Contextual associations represented both in neural networks and human behavior

Contextual associations facilitate object recognition in human vision. However, the role of context in artificial vision remains elusive as does the characteristics that humans use to define context. We investigated whether contextually related objects (bicycle-helmet) are represented more similarly in convolutional neural networks (CNNs) used for image understanding than unrelated objects (bicycle-fork). Stimuli were of objects against a white background and consisted of a diverse set of contexts (N=73). CNN representations of contextually related objects were more similar to one another than to unrelated objects across all CNN layers. Critically, the similarity found in CNNs correlated with human behavior across three experiments assessing contextual relatedness, emerging significant only in the later layers. The results demonstrate that context is inherently represented in CNNs as a result of object recognition training, and that the representation in the later layers of the network tap into the contextual regularities that predict human behavior.

Visual Cryptography Using Binary Amplitude-Only Holograms [Invited]

Visual cryptography (VC) is developed to be a promising approach to encoding secret information using pixel expansion rules. The useful information can be directly rendered based on human vision without the usage of decryption algorithms. However, many VC schemes cannot withstand occlusion attacks. In this paper, a new VC scheme is proposed using binary amplitude-only holograms (AOHs) generated by a modified Gerchberg-Saxton algorithm (MGSA). During the encryption, a secret image is divided into a group of unrecognizable and mutually-unrelated shares, and then the generated shares are further converted to binary AOHs using the MGSA. During image extraction, binary AOHs are logically superimposed to form a stacked hologram, and then the secret image can be extracted from the stacked hologram. Different from conventional VC schemes, the proposed VC scheme converts a secret image into binary AOHs. Due to the redundancy of the generated binary AOHs, the proposed method is numerically and experimentally verified to be feasible and effective, and possesses high robustness against occlusion attacks.

Foveal vision predictively sensitizes to defining features of eye movement targets

Despite the fovea's singular importance for active human vision, the impact of large eye movements on foveal processing remains elusive. Building on findings from passive fixation tasks, we hypothesized that during the preparation of rapid eye movements (saccades), foveal processing anticipates soon-to-be fixated visual features. Using a dynamic large-field noise paradigm, we indeed demonstrate that sensitivity for defining features of a saccade target is enhanced in the pre-saccadic center of gaze. Enhancement manifested in higher Hit Rates for foveal probes with target-congruent orientation, and a sensitization to incidental, target-like orientation information in foveally presented noise. Enhancement was spatially confined to the center of gaze and its immediate vicinity. We suggest a crucial contribution of foveal processing to trans-saccadic visual continuity which has previously been overlooked: Foveal processing of saccade targets commences before the movement is executed and thereby enables a seamless transition once the center of gaze reaches the target.

Human EEG recordings for 1,854 concepts presented in rapid serial visual presentation streams

The neural basis of object recognition and semantic knowledge has been extensively studied but the high dimensionality of object space makes it challenging to develop overarching theories on how the brain organises object knowledge. To help understand how the brain allows us to recognise, categorise, and represent objects and object categories, there is a growing interest in using large-scale image databases for neuroimaging experiments. In the current paper, we present THINGS-EEG, a dataset containing human electroencephalography responses from 50 subjects to 1,854 object concepts and 22,248 images in the THINGS stimulus set, a manually curated and high-quality image database that was specifically designed for studying human vision. The THINGS-EEG dataset provides neuroimaging recordings to a systematic collection of objects and concepts and can therefore support a wide array of research to understand visual object processing in the human brain.

Human Vision Reconstructs Time to Satisfy Causal Constraints

The goal of perception is to infer the most plausible source of sensory stimulation. Unisensory perception of temporal order, however, appears to require no inference, because the order of events can be uniquely determined from the order in which sensory signals arrive. Here, we demonstrate a novel perceptual illusion that casts doubt on this intuition: In three experiments ( N = 607), the experienced event timings were determined by causality in real time. Adult participants viewed a simple three-item sequence, ACB, which is typically remembered as ABC in line with principles of causality. When asked to indicate the time at which events B and C occurred, participants’ points of subjective simultaneity shifted so that the assumed cause B appeared earlier and the assumed effect C later, despite participants’ full attention and repeated viewings. This first demonstration of causality reversing perceived temporal order cannot be explained by postperceptual distortion, lapsed attention, or saccades.

Effects of color in humans generated by led lighting systems

This research evaluates the effects of color in humans generated by LED lighting systems. This evaluation allows us to know the different parameters that these systems generate, such as: the temperature and color of light suitable for human vision and thus be able to implement them in LED lighting designs for work areas. Having an adequate lighting system contributes to visual health and safety by avoiding work accidents. Each color has a different percentage of light reflection and in the same way each color affects us optically and psychologically by causing different sensations and perceptions in human beings such as visual fatigue, exhaustion, eye disorders, lacrimation, irritation, stress, migraines and even impaired vision Taking care of visual health is of vital importance. The methodology for evaluating lighting designs will be with NOM-030-ENER-2012. The contribution of this research will be to know the optimal colors and color temperature of light to use in LED lighting systems and to help lighting system designers in the selection of suitable LEDs and area colors that contribute to the visual health of humans

Fast 3D Content Update for Wide-Angle Holographic Near-Eye Display

Near-eye holographic displays are the holy grail of wear-on 3D display devices because they are intended to project realistic wide-angle virtual scenes with parameters matching human vision. One of the key features of a realistic perspective is the ability to move freely around the virtual scene. This can be achieved by addressing the display with wide-angle computer-generated holograms (CGHs) that enable continuous viewpoint change. However, to the best of our knowledge there is no technique able to generate these types of content. Thus, in this work we propose an accurate and non-paraxial hologram update method for wide-angle CGHs that supports continuous viewpoint change around the scene. This method is based on the assumption that with a small change in perspective, two consecutive holograms share overlapping data. This enables reusing the corresponding part of the information from the previous view, eliminating the need to generate an entirely new hologram. Holographic information for the next viewpoint is calculated in two steps: first, a tool approximating the Angular Spectrum Propagation is proposed to generate the hologram data from previous viewpoint; and second, the efficient Phase Added Stereogram algorithm is utilized for generating the missing hologram content. This methodology offers fast and accurate calculations at the same time. Numerical and optical experiments are carried out to support the results of the proposed method.

Review of Visual Saliency Prediction: Development Process from Neurobiological Basis to Deep Models

The human attention mechanism can be understood and simulated by closely associating the saliency prediction task to neuroscience and psychology. Furthermore, saliency prediction is widely used in computer vision and interdisciplinary subjects. In recent years, with the rapid development of deep learning, deep models have made amazing achievements in saliency prediction. Deep learning models can automatically learn features, thus solving many drawbacks of the classic models, such as handcrafted features and task settings, among others. Nevertheless, the deep models still have some limitations, for example in tasks involving multi-modality and semantic understanding. This study focuses on summarizing the relevant achievements in the field of saliency prediction, including the early neurological and psychological mechanisms and the guiding role of classic models, followed by the development process and data comparison of classic and deep saliency prediction models. This study also discusses the relationship between the model and human vision, as well as the factors that cause the semantic gaps, the influences of attention in cognitive research, the limitations of the saliency model, and the emerging applications, to provide new saliency predictions for follow-up work and the necessary help and advice.

Performance and stability of fly-inspired, saccade-based hybrid control

In a way analogous to human vision, the fruit fly Drosophila melanogaster and many other flying insects generate smooth and saccadic movements to stabilize and shift their gaze in flight, respectively. It has been hypothesized that this combination of continuous and discrete movements benefits both flight stability and performance, particularly at high frequencies or speeds. Here we develop a hybrid control system model to explore the effects of saccades on the yaw stabilization reflex of D. melanogaster. Inspired from experimental data, the model includes a first order plant, a Proportional-Integral (PI) continuous controller, and a saccadic reset system that fires based on the integrated error of the continuous controller. We explore the gain, delay and switching threshold parameter space to quantify the optimum regions for stability and performance. We show that the addition of saccades to a continuous controller provides benefits to both stability and performance across a range of frequencies. Our model suggests that Drosophila operate near its optimal switching threshold for its experimental gain set. We also show that based on experimental data, D. melanogaster operates in a region that trades off performance and stability. This trade-off increases flight robustness to compensate for environmental uncertainties such as wing damage.

Characterization of retinal development in 13-lined ground squirrels

Purpose: The cone-dominant, 13-lined ground squirrel (13-LGS) retina mimics the human foveal region but retinal development in this useful rodent species has not been reported. Here, the embryonic and postnatal development of the 13-LGS retina was studied to further characterize the species as a practical alternative animal model for investigating cone-based vision in health and disease. Methods: The spatiotemporal expression of key progenitor and cell type markers was examined in retinas from defined embryonic and postnatal stages using immunohistochemistry. Changes in the postnatal gene expression were also assessed by qPCR. Results: The 13-LGS neuroblastic layer expressed key progenitor markers (Sox2, Vsx2, Pax6, and Lhx2) at E18. Sequential cell fate determination evidenced by the first appearance of cell type-specific marker labeling was: at E18, ganglion cells (Brn-3A, HuC/D) and microglia (Iba1); at E24-25.5 shortly before birth, photoreceptor progenitor (Otx2, Recoverin), horizontal and amacrine cells (Lhx1, Oc1); and at P15, bipolar cells (Vsx1, CaBP5) and Muller glia cells (GS, Rlbp1). Photoreceptor maturation indicated by opsin+ outer segments and PNA labeling of cone sheaths was completed at the time of eye opening, P21-24. Conclusions: The timeline and order of retinal cell development in the 13-LGS generally matches that recorded from other mammalian models but with a stark variation in the proportion of various cell types due to cone-dense photoreceptors. This provides a baseline for future examinations of developmental, disease model, and stem cell approach studies employing this emerging rodent model of human vision.