Cognitive: Week 2: Percieving Objects

Recognising a thought can be done by:

1) Detecting features (Bottom-up Processing) - this is when we encounter an object and we see it from different viewpoints. So, we don't just detect features because an objects context may not make sense.

2) Interpreting input: We rely on the context and use Top-down processing and Bottom-up Processing to make sense of the context. 

Gestalt psychology is concerned with perceptual organisation, it describes how we link or separate objects, thus it is our guiding principles (Pragnanz). 

There are 5 Pragnanz:

 1) Similarity,     (2) Proximity,     (3) Good continuation,     (4) Closure,     (5) Simplicity.

Gestalt particularly intrigues in Figure-ground segregation: separating an object from its background.

Strengths of Gestalt: It focusses on fundamental issues such as reconising objects in everyday life as iportant.

Weaknesses of Gestalt: Ignores prior knowlege and learning, making it an REDUCTIONIST approach. 

1) Visual Search : This takes longer when searching in a combination.

2) Theory of Object Perception: This involvevs feature nets which can use bottom-up processing or top-down processing and is a hierarchy of detectors in order to recognise simple stimuli. 

From The theory of object perception, Biederman suggested that all items can be reduced to geons ( Geometric ions).

- We often percieve constant objects despite context, this is due to elements of perceptual constancy:

A) What an object is:

1) Size constancy,      (2) Shape Constancy,     (3) Colour Constancy 

B) Where an object is via Depth cues:

4) Bioncular cues,   (5) Oculomotor cues,   (6) Monocular (pictorial) cues,  (7) Motion cues.

- Face recognition involves holistic processing (the integration of an entire object (face).

- There is evidence for holistic processing:

         1) Face inversion effect

         2) Prosopagnosia (face blindness)

- Theory of face recognition by Bruce and Young 

       > Arguably the most influential theory.

Face   > Structural Encoding > 1) Face Memory or 2) Expression Analysis > Person identification > Name generation.