Waving-hand steganography by use of 480-fps LED panel

Hirotsugu Yamamoto, Farhan Syahmi and Shiro Suyama

Decoding of encrypted image provides fun. This sensation is evoked when unconscious information turns apparent. In this research we aim to propose a new way of displaying information or messages so that watching digital signage will not only receive information but will become an enjoyable experience to viewers. A text is hidden in successive frames so that it is unnoticeable when being displayed on an LED panel with a high frame rate (HFR). This hidden text is decodable by viewing through a waving hand. In order to implement a new technique of providing secret information, we have developed a high-frame-rate LED display. Full-color images are refreshed at 480 frames per second. A secret text was embedded into HFR images and represented at 240 fps. When observations were done without waving hand, only white screen or background image can be seen. Secret message was successfully embedded into the background image. However, the secret message was decoded when waving hand. Decoding of the secret message without any special gear has been successfully realized. Observers can decode the hidden image with naked eye by waving hand in front of their eyes.

Local contextual interactions result in global shape misperception

David Badcock, Edwin Dickinson, Clare Harman and Renita Almeida

When required to discriminate between similar shapes the visual system often sacrifices accuracy of representation while exaggerating difference. Adaptation to a stimulus, often results in properties of subsequently presented stimuli being repelled along identifiable axes. Adaptation to a path deformed from circular by a sinusoidal modulation of radius results in a circle appearing to have modulation in opposite phase. We show that the tilt aftereffect (TAE) can account for this. Deformation is also observed when only the orientations of local elements are modulated. In contrast to the TAE, this modification reduces the perceived angle between the local orientation cues. We show that this effect is consistent with a known illusion with a local cause, the Fraser illusion. Both effects can be accommodated by a population of orientation tuned neuron encoding local orientations as a vector sum of contextually modified activation. It is proposed that shape aftereffects, which can be predicted using a spatially continuous representation of the TAE in a TAE field, might serve to de-correlate the responses of higher level shape analyzers with the Fraser illusion being a consequence of using collinear orientation to bind a disrupted path. These distinct processes can arise from the same neural architecture.