Effects of Corporate Restructuring - Essay Example

Comparison between Viewpoint-Invariant Theories and the Viewpoint-Dependent Theories Number: Introduction The way humanbeings view objects is determined by a number of factors as explained by the two popular theories of object perception namely the Viewpoint-Invariant Theories and the Viewpoint-Dependent theories. Object perception, emotions, and behaviors among other processes are determined by various factors that are evaluated in a branch of psychology known as cognitive psychology. Cognitive psychology also explains mental perspectives that deal with psychological processes like problem solving, conceptualization, thinking, judgment, attention, remembering, memory, perceiving, reasoning, believing among other key aspects that human beings encounter in their daily activities (Quinlan & Dyson, 2008). It explains why different people have different views over a similar issue. However, this essay is going to focus on the element of perception by comparing and contrasting the two theories applied in explaining object perception, while at the same time drawing concise illustrations where necessary. The essay will focus on the Recognition-by-components theory developed by Irving Biederman as an example of Viewpoint-Invariant theory and Gibson’s Theory of Direct Perception to represent Viewpoint-Dependent theories in order to make a clear understanding between these two types of theories First, it is essential to look at the overview of the two theories. Recognition-by-components theory (RBC) was developed by Irving Bierdman in 1987, purposely to explain the element of object recognition by different people. The theory explains human ability to identify objects irrespective of variations such as change in size and orientation of the image. It also explains how distorted images of objects are effectively identified by human visual system. The theory suggests objects’ illustrations are kept in the brains inform of structural descriptions with each structural description comprising of specifications of the objects geons (geometricions), as well as the geons interrelationships (Quinlan & Dyson, 2008). Then, the visual system analysis the perceived object by explaining it into its fundamental geons, a step that is followed by determination of the interrelationship of the object. This process involves establishment of aspects like the position or the location of the object as well as its size. Then, the interrelationships and the geons of the object being observed are harmonized against the structural portrayals stockpiled in the brain (Pashler, & Yantis, 2002). The effectiveness of the processes explained above is the key determinant of successfulness of object’s identification. If the geons, structural imageries, and the interrelationships match, then the item is recognized successfully. On the other hand, the observer cannot identify the object if there is a mismatch or the brains have not stored images of similar geons. This theory’s suggestions over object recognition are the same as those of speech recognition whereby tiny set of phonemics are joined using organizational guidelines to release millions of various words. According to the theory, geons works as phonemics while the 3-D interrelationship act as organizational rules. It is estimated that a small number of geons like 36 are capable of producing millions of different objects (Pizlo, 2010). This theory was developed purposely to explain primal identification of objects (Sokhanvar, 2013). This process (primal recognition) does not use higher-level cognitive procedures such as shading, coloring, texturing among others in differentiation of objects. A bottom-up procedure helps in objects’ recognition. In simple terms, the theory assumes that people are able to see objects by separating them into their main constituent parts referred to as geons. According to theorist Biederman, geons are based on three dimension (3-D) shapes like cones and spheres among others that can be organized in various ways to form an almost unlimited quantity of objects. An extra top down processing of objects may arise due application of environmental cues like context in identifying difficult instances of objects. For example, it would be easier to identify a pen placed on a piles of books rather than when placed in a pile of woods or leaves (Pizlo, 2010). The assumption that geons can be produced from the contrast of five detectable features of edges in a 2-D image can be termed as the most fundamental presumption made by the theory. The five properties in this case include the curvature (different parts of the curve), parallel lines, co-termination (the point where at least two points meet hence stopping from any continuation), co-linearity which refers to points that branch from the main line as well as symmetry and asymmetry. The detection of these properties does not vary with changes in the position of viewing or that of the quality of the image (Sokhanvar, 2013). This gives room for an easy object perception especially if the image of the object is projected from a unique point or is distorted. Therefore, this theory explains the relationship between the typical philosophies of pattern recognition and the perceptual organization. Object recognition can therefore be attributed to the principle of componential recovery (Pizlo, 2010). At the same time, objects can easily be recognized as long as the arrangement of the geons that are more than two is recoverable. Such recognition would be easy despite the fact that the object is distorted, degraded, rotated, or blocked. Experiments made on human observers about object perception attest the theory whereby an object can project unlimited image configurations to the retina. The position of the viewer to the object can change consecutively with each locational change leading to a different 2-D projection (Pizlo, 2010). This would happen even if the object to be recognized is obstructed by other objects like vegetation. Moreover, the object does not have to be presented as a colored or highly textured image, but as easy illustrations. The object could also be missing some of its constituents and still be easily recognized. Some of the highlights in the Recognition-by-component theory are the claims that human beings are able to see objects as a result of combination of geons, which Beardman claim that their number is less than 36 (Pashler, & Yantis, 2002). According to Beardman argument, in order for one to identify a cup, the cup is broken into two of its major parts, namely the handle and the cylinder and he continues to argue that the same thing applies even for identification of bigger objects. The only difference that occurs is that larger objects have a relatively higher number of constituent parts (geons). Then, our minds compare the perceived geons with the objects’ images stored in the brains in order to identify that particular object (Pizlo, 2010). Another unique claim by the theory is that human beings identify objects by focusing on their edges and concavities. By looking at the edges, Biederman argues that humans are able to maintain a similar view of the object irrespective of the position or angle at which a person is looking at the object, known as viewpoint invariance. He articulates this to the fact that edge properties of geons never change. He also argues that by looking on the objects concavities (the intersection of two edges), people are able to identify the difference between various geons (object parts). Gibson’s Theory of Direct Perception commonly referred as Ecological theory is a type of a Viewpoint-Dependent Theory (Quinlan & Dyson, 2008). The theory is a bottom-up theory that suggests that change of the observer’s position is a vital element in identifying the object. It suggests that identification of objects depends on previous perceptions with the angle and size of the object being observed remaining constant. Variations such as size and observation angles would automatically make the observer not recognize the object. The theory proposes that there exists no need of processing the images people see. For a theory to be upward, it means that information is examined in a single direction; from simple data analysis to complex data analysis. The theory suggests that processing of the rental image is the first step in recognizing objects. This is followed by the description of the object’s images as the second step with structural description as the third step. This involves the formation of 3-D images in the brains for object identification. Gibson’s theory on direct insight focuses on the overall fast and accurate observation on the environment. Nevertheless, the theory does not provide explanation on the reasons for inaccuracies such as illusion. The theory cannot explain the occurrence of perceptual inaccuracies such as inclination by human beings overestimating vertical scopes relatively to the horizontal ones (Quinlan & Dyson, 2008). Some of the most obvious differences between the two types of theories (Viewpoint-Invariant Theories and the Viewpoint-Dependent theories) include. First, Viewpoint-Invariant Theories claim that recognition of the object is based on structural information like constituent parts, which allows recognition of the object irrespective of the observer’s viewpoint or his or her observation angle. This means that objects are recognizable from any viewpoint (Goldstein, 2010). On the other hand, Viewpoint-Dependent Theories’ claims contravene the above argument by suggesting that object’s recognition is determined by the position or location of the observer as well as the observation angle. This means that the speed of recognition and the accuracy relies on the angle and the position of the viewer. The theory has three major elements namely optic flow pattern, invariant features as well as affordances (Pashler & Yantis, 2002). Another key distinction between the two theories is the fact that viewpoint-invariant theories claim that human beings do not require much memory to recognize the same object from different viewpoints. According to the theories, this result from the fact that same object can be identified from any viewing angle as explained above, hence little memory needed to store the illustrations of the object. A larger memory on the other hand is needed for Viewpoint-Dependent Theories for storing different illustrations of the object since the observer has to master every viewpoint from all angles. In addition, according to Viewpoint-Dependent Theories, the accuracy of the observer in identifying the object relies on whether the observer is familiar with that particular object placed in that angle or not. On the other hand, viewpoint-invariant theories suggest that the observer can identify the object from any angle and does not have to cram all the object’s images from all angles (Goldstein, 2010). One of the key similarity between the two types of theories is that recognition of the object relies on whether the observers mind has stored images of an exactly object or that is closely similar. This means that it is almost impossible for an individual to identify items that he or she has never seen before. At the same time, both theories apply the bottom up approach in processing information. This means that brains start by processing the easy information before embarking on the complex one (Pashler & Yantis, 2002). In conclusion, perception of objects significantly relies on whether the observer’s brains have stored similar images as the ones portrayed by the object being observed. Identification of objects is therefore claimed to be a bottom-up process that involves processing of the received information from the simplest data to the most complex. This is suggested by the two types of theories of perception namely, the Viewpoint-invariant and the Viewpoint-dependent theories. These two types of theories have a number of differences as even though they share some few common similarities. Some of the major differences between the two theories are the issue of change in position and object size. Viewpoint-invariant theories claim that such changes cannot affect observer’s ability to recognize objects while viewpoint-independent theories on the other hand claim that they such changes do affect observer’s ability to identify objects. Other common differences include the issue of memory used for storing images with viewpoint invariant requiring lesser memory than viewpoint-dependent. The most obvious similarity between these two types of theories is the fact that they both recognize the bottom-up as the process of processing the received information. The two theories also argue that identification of an object relies on whether the observer’s brains have stored images of a similar object or that closely looks like the one being observed. With this, it appears as if human beings apply both theories in recognizing objects. References Goldstein, E. B. (2010). Sensation and perception. Belmont, Calif: Thomson Wadsworth. Pashler, H., & Yantis, S. (2002). Sensation and perception. New York, NY: Wiley. Pizlo, Z. (2010). 3D shape: Its unique place in visual perception. Cambridge, Mass: MIT Press. Quinlan, P. T., & Dyson, B. (2008). Cognitive psychology. Indianapolis, Ind: Prentice Hall. Sokhanvar, S. S. (2013). Tactile sensing and displays: Haptic feedback for minimally invasive surgery and robotics. Chichester, West Sussex, U.K: John Wiley & Sons. Read More