PG KZVI cv6

When we open our eyes, it seems to us that we are aware of ordinary objects and events such as tomatoes, elephants, and bouncing balls. Direct perception means that we are indeed aware of such environmental objects, without mediating representations or inference processes. Yet this appears to present a paradox, for we are not in physical contact with environmental objects, but only with the light that arrives at our receptors, the neural signals it elicits, or resulting sensory states~\cite{gold}. This paradox can be observed in desert, where the refraction of light causes phenomena usually known as mirage.

Something very similar is happening while displaying the computer graphics results to observer. Computer produces for observer only clusters of coloured dots and the observer very easy supposes that the rendered result is real scene and he is capable of immersion and presence.

Two visual systems

One system, vision-for-perception, allows us to recognize objects and their relationships, enabling us to build up a knowledge base about the world. This is the system we are more familiar with, the one that gives us our conscious visual experience - and allows us to see and appreciate objects in the world beyond our bodies. The other system, vision-for-action, provides the visual control we need to move about and interact with those objects.

The ventral visual stream, which projects to higher-order visual areas in the ventral part of the temporal lobe, mediates vision-for-perception. The dorsal visual stream, which projects to visuomotor areas in the posterior parietal lobe, mediates vision- for- action. Importantly, the dorsal stream also gets visual inputs directly from subcortical visual structures that bypass the early visual areas in the cortex.

The ventral stream, which plays the critical role in vision-for-perception, transforms incoming visual information into perceptual representations that embody the enduring characteristics of objects and their relations. These representations allow us to perceive the world beyond our bodies, to share that experience with other members of our species, and to plan a vast range of different actions with respect to objects and events that we have identified. The perceptual machinery in the ventral stream that has evolved to do this is not linked directly to specific motor outputs, but instead accesses action plans via cognitive systems that rely on memory, semantics, spatial reasoning, and communication with others.

But even though the two systems transform visual information in quite different ways, they work together in the production of adaptive behaviour. In general terms, one could say that the selection of appropriate goal objects depends on the perceptual machinery of the ventral stream, whereas the visual control of the goal-directed action is carried out by dedicated online control systems in the dorsal stream ~\cite{gold}.

Perception of depth – stereopsy

Depth perception in general can be understood as a reconstructive process that interprets the retinal image in our eye such that a 3D object arises in our mind. Pictures and films can also provide vivid impressions of depth. The ability of observers to perceive the 3D shapes~\cite{shaperec} of objects from patterns of light that project onto the retina surprises scientist of many different disciplines (psychology, neuroscience, computer science, physics and mathematics) for almost 2 millennia and it is still an active area of research. If not for own experience, we could conclude that the visual perception of 3D shape is a mathematical impossibility.

This pictorial depth differs in nature. It is a constructive process of its own and presents an additional level of difficulty. Normal vision allows us to glean information about an object's shape and color as well as about such things as its spatial relations, its mass, and its potential danger. Normal vision typically reconstructs the real world object which gives rise to the retinal image with admirable precision. This is possible because our visual system is able to resolve the many ambiguities present in the retinal image. Pictorial depth is both more confined and broader than normal depth.

Figure \ref{Figure2.2.3} illustrates the nature of the (re)constructive processes in pictorial viewing compared with normal viewing. In normal viewing, a large number of 3D objects would qualify as permissible reconstructions that could be made on the basis of one given retinal image~\cite{gold}.

In the praxis the depth perception means that if we focal our eyes on some random point (in Figure \ref{Figure2.2.2}), this flashbacks always in the middle of retina of every eye, where the biggest density of photoreceptors is. Other visible points are flashbacked to retina in the distance corresponding to distances from the fixed point.

The singleness principle One object can't be on two places at the same time. It means that we can couple point in the picture from left eye exactly with the one point seeing with the right eye.

Coherence principle Seeing that the surfaces of objects are in most times non-transparent and smooth, the neighbour points in the picture usually corresponds with points in approximately the same depth (distance from eye) ~\cite{takac}.

How we perceive 3D shape with a 2D retina is often considered a classic example of under-determined perception. But Gibson argued that such constant properties as surface shape are specified by constant spatio-temporal patterns of stimulation, which he called higher-order invariants. Higher-order patterns of texture, shading, optic flow, and binocular disparity indeed specify the 3D shape of a smooth surface. However, this information is not sufficient to determine Euclidean shape (absolute depth, slant, curvature) as researchers have typically assumed, but rather qualitative surface shape (hills, dales, ridges, valleys, and plains)~\cite{gold}.

James Gibson~\cite{gibsonperception} thought that to perceive a physical environment, it must have a one-to-one correspondence with some measurable optical stimulation. According to this view, the problem of 3D shape perception is to invert (or partially invert) a function of the following form:

where $ \phi $ is the space of environmental properties that can be perceived, and $ \Lambda $ is the space of measurable image properties that provide the relevant optical information. The primary difficulty is that the relation between $ \phi $ and $ \Lambda $ is almost always a many-to-one mapping in natural contexts (for any pattern of optical stimulation, there is infinity of possible 3D structures that could have produced it).

The inherent ambiguity of visual information is not always a serious problem as it might appear at first. Recent theoretical analyses have shown, for example, that the animation around an object (the change of saliency part of object and its shadow) could fill in the information gaps. It is important to note that these are linear transformations. Because motion is such powerful source of information, especially when presented in combination with stereo vision~\cite{stereoandmotion}, it should not be surprising that they are primary for the perception of 3D shape in natural vision.

By motion and stereovision could be also properties of material~\cite{perceptioncg} of virtual or real object such as translucency, glossiness and texture can be easily detected. It depends pretty much on the way how light interacts with different materials. We can easily say if the object is made of wood or marble, if it is liquid or solid or if it is undulated or smooth.

Spatial relations

Spatial relations~\cite{qvort} between objects are the basis for their possible grouping and clustering, but eventually the spatial relation between the objects and the observer determines how the objects' spatial distribution is perceived.

Viewpoint Dependency. The visual experience of seeing objects is very dependent on the viewpoint and the view direction of the observer relative to the objects under observation. The overall spatial distribution will appear different from different viewpoints and even some object properties like shape and pose are intrinsically viewpoint dependent.

All this indicates that checking all the visual potential of the visual world means exhaustive search through all view directions from all possible viewpoints - a rather impossible task. It also indicates the importance of convenient facilities for navigation in visual space and that these facilities should be used in connection with appropriate search strategies inspired by the human talents for serendipitous search.

Viewpoint distance (the distance from the observer) to the objects observed is varying as the observer "moves" around in the virtual world. This distance is a very important factor, when evaluating how an observer perceives the alternative object properties. As viewpoint distance increases, the significance of the various object properties may change and most of them will be reduced and eventually vanish.

How to trick perception and catch the eye

Another possibility of how to attract observers attention is to make something unusual while shooting the animation in the virtual museum.

Visual illusions don't correspond to the physical reality around an observer. They are highly subjective percepts. The terms visual illusion and optical illusion are often used interchangeably. However, optical illusions do not result from brain processes. Experiencing a visual illusion, we may see something non-existing, we may not see something obvious, or see something different from what is there. Using visual illusions could stimulate observers mind to concentrate better on shown visitation. And it demonstrates how the brain fails in recreating the physical world.

An example of a world-known visual illusion is the Ebbinghaus illusion, named for its creator, Hermann Ebbinghaus. It is based on the fact that two identical circles seems dissimilar when they are placed side by side, one surrounded by large circles and the other surrounded by small circles. (in Figure \ref{Figure5.2}).

The Ebbinghaus illusion cannot be explained by the physical properties of the visual stimulus (such as the optical illusions can be). It is due to neural processes that the eye includes the surrounding of examined object and compares it with its context. Visual artists have often used their insights regarding perception to create visual illusions in their artwork. Historically, first visual illusions (we don't consider them as illusions) are paintings - artists had devised a series of techniques to "trick" the brain. We don't have a problem to think that displayed object is 3D a series of brushstrokes was in fact a still life~\cite{gold}.

Visual illusions are developed either wittingly – we apply known visual principles to stimulate observer and experiment with variations illusions, or we discover the completely accidentally. Most of these accidental illusions were spotted in the nature first and the artist tried to understand and replicate the conditions leading to the unusual percept and of course illusions may be discovered through the application of known physiological principles of visual processing in the brain~\cite{gold}.

Categories of Visual Illusions

There have always been efforts to classify visual illusions into categories with varying degrees of success. One of more obstacles is that some visual illusions that seem similar may be due to disparate neural processes, whereas other visual illusions that are phenomenologically different may be related at a neural level~\cite{gold}. . However, the neural underpinnings of many visual illusions - especially those discovered recently - are not understood some representative categories of visual illusions were built-up.

• Adaptation Illusions This visual illusion was described in Aristotle’s Parva Naturalia. It is mostly known as “waterfall illusion,” and it can be observed while looking at flowing water. Watching the flowing water for a while, and then quickly shifting center of gaze to the stationary objects next to the water (for instance, the rocks) will trick the mind. The stationary objects will appear to flow in the opposite direction to that of the water. The illusion occurs because neurons that detect motion in a specific direction become adapted in response to steady stimulation.
• Brightness Illusions Brightness and color illusions often occur because the brain does not directly perceive the actual wavelength and light reflected from objects in the world. For instance, the same gray square (in Figure \ref{illusionscat} b) will look lighter when surrounded by black than when surrounded by white.
• Color Illusions Some classical color illusions are based on simultaneous color contrast. For instance, a gray circle will take on a reddish hue when placed against a green background, and a greenish hue when placed against a red background. The underlying neural processes are not well understood, but current theories point toward retinal circuits.
• Illusions of Size The apparent size of an object is changed, usually due to surroundings. In the Ponzo illusion (in Figure \ref{illusionscat} a), two horizontal lines of the same length are placed on a pair of converging lines (it evocates train tracks). The upper line seems longer than the lower line. The illusion is probably due to the fact that the brain interprets the upper line as farther away than the lower line. The Ebbinghaus illusion, discussed in the introduction to this entry, is another example of a classic size illusion.
• Shape and Orientation Illusions In this kind of illusions an object appears to take on shapes or orientations different from the actual physical ones. Distortion effects are often produced by the interaction between the observed object and and the shapes or orientations of other nearby figures. A classical example is the Café Wall illusion (in Figure \ref{illusionscat} c). The black and white tiles in the Café Wall are perfectly straight, but look tilted.

• Invisibility Illusions In an invisibility illusion, observers fail to perceive an existing object in the scene. In motion-induced blindness, he fixates the center of a display consisting of several stationary circles and a surrounding cloud of moving dots.
• Illusory Motion Some patterns (still and repetitive) can create the illusory of motion. The effect is stronger if the observer moves eyes around the figure. If he keeps eyes still, the illusion tends to weaken or even disappear. For instance, in the Rotating Snakes illusion created by Akiyoshi Kitaoka, the “snakes” appear to rotate. (in Figure \ref{Figure5.44})
• Stereo-Depth Illusions The left eye and the right eye of observer convey different views of the world to his brain. The brain integrates these two images into a single stereo image, which conveys a sense of depth. This is in the fact the principle behind stereo-depth illusions. The classic example is wallpaper illusion (in Figure \ref{illusionscat} d), which arises when observing a pattern of horizontal repetitive elements, such as in wallpaper.

Camera angle

Camera angles can deliver important information, create impact on audience, facilitate editing, it can emphasize and determine a point, camera angles are a subtle yet vital devise in film making.

With the right combination of camera movements and viewpoint angles \cite{filmshot} we don't need verbal expression for describing the situation on the scene. The relationship between the camera and the object being filmed gives emotional information to observer, and guides their judgement about the character or object in shot. The more extreme the angle is the more symbolic should be the expression of the shot.

• The Bird's-Eye view. This shows a scene from directly overhead, a very unnatural and strange angle.
• Extreme close up shows very little background, and concentrates on either a face, or a specific detail of scene.
• Medium Shot shoots a figure from the knees/waist up and is normally used for dialogue scenes, or to show some detail of action.
• Low angle. These increase height (useful for short actors like Tom Cruise or James McAvoy) and give a sense of speeded motion.
• High angle The camera is elevated above the action using a crane to give a general overview.
• Aerial shots are usually done with a crane or with a camera attached to a special helicopter to view large landscapes.
• Over the shoulder (or third-person shot) is a shot of someone or something taken from the perspective or camera angle from the shoulder of another person.
• Establishing Shot sets up, or establishes the context for a scene by showing the relationship between its important figures and objects.
• The Eye-Level shot is the most common angle seen in movies. Scenes are shot at roughly the same level as an observer would see the scene.
• Oblique angle is shot by literally tilting the camera frame. It can be used to suggest a sense of “crookedness” and anxiety, or, in the case of some television news shows and music video programs, a sense of playfulness

The usage of different camera angles by shooting while the virtual museum exploration could increase not only the attention of observer - with the right angle we can evoke the atmosphere of examined artpiece, but we can improve the visit regarding to the amount of information.

Kuleshov effect

The Kuleshov effect takes its name from Lev Kuleshov, an influential filmmaker in the mid-twentieth century Soviet Union, who illustrated it (in Figure \ref{Figure5.1}). It is based on the fact that human mind is accustomed of supplementing the information. Kuleshov shot in this experiment a single long close-up of an actor sitting still without any face expression. He then intercut this shot with various different shots (comprised a bowl of soup, a woman in a coffin, and a child with a toy bear).

The audience was surprised of the sensitivity of the actor's range after preview of final movie although there was no visible difference between various frames of actor. The mind of observer simply adapted to the film as a whole. The viewer is presented with a situation or environment along with the academic fact that someone is experiencing it. He cannot simply accept the actor's evident emotion, as none is given, so he decides what the appropriate response would be and assigns it to the actor~\cite{kulesh}.

Gestalt psychology

• Similarity. It occurs when objects look similar to one another. Similarity seduces the observer to group object in the scene.

  

• Continuation. It occurs when the eye is compelled to move through one object and continue to another object.

  

• Closure. It occurs when an object is incomplete or a space is not completely enclosed. Our visual system is adapted to predict the shape of object by filling in the missing information.

  

• Proximity. It occurs when elements are placed close together. In spite of they are separated they location in the space makes observer think of them as a group.

  

• Figure and Ground. The eye differentiates an object form its surrounding area. The figure is in this case our examined object and Ground is the background.