{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

Pattern Recognition adv. (studt)

Pattern Recognition adv. (studt) - PERCEPTION PATTERN...

Info iconThis preview shows pages 1–8. Sign up to view the full content.

View Full Document Right Arrow Icon
PERCEPTION & PATTERN RECOGNITION Sense Modalities : Vision** Audition Haptic Smell Taste Sensation-detection Perception-identifying, recognizing I. Visual Perception/Low Level Vision - Distal stimulus : objects or targets in the real world - Proximal stimulus : a stimulus as it impinges on a sensory organ (e.g., 2D image on retina) A. Marr’s (1982) Computational Theory: Low level vision detects physical properties of the environment: surface, boundaries, depth, and motion. Things needed to program a Robot: Stored Representation of each object, sensory apparatus(color, shape, light/contrast), match up sensory data-memory, separate the object from its background, depth, movement. 1. Four stages of visual processing : a. Grey-level description : represents the intensity of the light energy at each point on the retinal image. This breaks the image into basic areas, and defines their boundaries.
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
b. Primal Sketch : Surface marks, object boundaries, shadows, and so on are derived from the grey-level image. Potentially significant regions are identified, and information about how they fit together is derived. c. 2 1 /2-D sketch : Orientation and depth of visible surfaces is made explicit; only includes visible parts of scene/object and thus is "viewpoint-centered". Thus, representation of the object is dependent on the angle from which it is being perceived. d. 3-D model representation (pattern recognition stage) Depicts the spatial organization of objects/shapes, but independently of their position on the retina (viewpoint independent). --basic units for object recognition are cylinders (Marr and Nishishara, 1978)
Background image of page 2
--decomposition of forms into their constituent cylinders, then matched against models stored in memory.
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
2. Modular elements : Surface, Color, Orientation, Texture (also Binocular vision, Motion), used to construct 2-1/2 D sketch, in which depth, color, etc. have been made explicit . B. Edge Extraction: Feature Analysis 1. Claims : --stimuli broken down into sets of elemental features --patterns distinguished from each other based on distinctive features (e.g., P vs. R) (Gibson)
Background image of page 4
2. Evidence a. Letters presented at fast presentation rates confused with visually-similar letters e.g. 29 errors on G: 21 as C, 6 as O, 1 as B, 1 as 9 (Kinney et al., 1966) b. Neisser’s Search for Z : RT longer for Angular Letter matrix because of feature competition
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Background image of page 6
c. Hubel & Wiesel : feature detectors (simple cells) found in primary visual cortex, which are sensitive to lines at a given angle in a
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 8
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}