Docstoc

HUL 211 Bottom up processing (PDF)

Document Sample
HUL 211 Bottom up processing (PDF) Powered By Docstoc
					Bottom up processing




    Snehlata Jaswal


     HUL 211 OBJECT PERCEPTION AND MEMORY
            Bottom up processing

Bottom-up processing is analysis that begins with the sense
receptors in the body and works up to the brain's integration of
sensory information. This form of processing suggests that
information is received in small units, and built into larger units
that carry meaning. This is the classic example of perception being
brought on through sensation. For example, if you are reading a
book, you experience (sense) the letters one at a time, and then
organize the letters into words, the words into phrases or clauses,
then sentences.




                     HUL 211 OBJECT PERCEPTION AND MEMORY
  Gibson (1966 ): Direct Perception
The theory of direct perception, by Gibson (1966), states that all
one needs for perception to occur is a visual system and an
environment in which to operate. All the cues necessary for
perception to take place, according to Gibson's theory, are already
present in the environment, and it is not necessary to invoke top-
down processing.




                     HUL 211 OBJECT PERCEPTION AND MEMORY
    Bottom up processing in vision
Two bases for bottom up processing of information,
resulting in binding:

Conjunctively coding neurons - Hubel and Wiesel (1959, 1962,
1968)

Neural synchrony – Gray, Engel, Konig, and Singer (1992), Singer
and Gray (1995)




                     HUL 211 OBJECT PERCEPTION AND MEMORY
                 Empirical evidence
Differential processing of features:

This refers to the fact that different features of an object are initially
processed in parallel, and to different time scales.

Numerous studies have demonstrated that various dimensions of
the stimuli are processed asynchronously by the visual system
(Aymoz & Viviani, 2004; Bedell, Chung, Ogmen, & Patel, 2003;
Lamberts, 2002; Moutoussis & Zeki, 1997; Nishida & Johnston,
2002, Viviani & Aymoz, 2001, Zeki, Watson, Lueck, Friston, Kennard,
& Frackowiak, 1991). Between colour and shape, it is debatable
which is perceived first, and the answer depends on the kind of
shapes used, but both are definitely processed much faster than
movement.
                        HUL 211 OBJECT PERCEPTION AND MEMORY
                Empirical evidence
Differential processing of features:
Psychophysical investigations into memory for features also suggest
that information in VSTM is encoded in many different parallel
channels (Magnussen, 2000; Magnussen & Greenlee, 1997;
Magnussen, Greenlee, & Thomas, 1996). Detection and
discrimination is easier across channels than within a channel.
Multiple decisions required by increasing set size are also more
difficult within a channel, though it is difficult to explain the step
change in memory when set size is four vs. when it is more than
four. Dimension weighting account (Muller & Krummenacher,
2006a, 2006b) suggests that features of the objects are assigned
different weights, and guide attention to the objects. E.g., location
is a more powerful guide than form, which is a more powerful guide
than colour and so on. According to this account, features of
different objects ‘pull’ attention to the object.
                      HUL 211 OBJECT PERCEPTION AND MEMORY
                Empirical evidence
Differential processing of features IN MEMORY:

Lamberts and Kent (2008) investigated the time course of
perception and recognition of three features, colour, shape, and
orientation, as combined in either one or two study stimuli. Their
work confirmed that perceptual processing of features happens at
different rates, colour being processed the fastest, followed by
shape, followed by orientation. Using a recognition paradigm akin
to change detection, they also showed that the time course of
retrieval of features mirrored that of perception, and thus differed
for each feature, but only when the memory load exceeded
capacity at six features (defining two different objects) to be
remembered. There were no differences when only three features
(defining one object) were to be remembered.
                      HUL 211 OBJECT PERCEPTION AND MEMORY
                 Empirical evidence
Attention capture
Theeuwes(1992, 2004); Theeuwes, Reimann, & Mortier (2006) use a
singleton distracter defined by a different dimension than the one
defining the singleton target in their experiments. Theeuwes (1992)
used a distracter defined by colour (the only red among all green),
and a target defined by shape (the only diamond among circles). The
initial check confirmed that RTs were quicker to colour than to shape,
showing it to be more salient. Then participants performed under
two conditions, one in which the distracter was present, and the
other in which it was not. Results showed significant distracter
interference, in that RTs to the target were significantly slower when
the distracter was present.

The embarrassing question for adherents of top-down influences was
why participants were unable to ignore the distracter; despite the fact
that they knew the dimensions defining the target and the distracters.
                       HUL 211 OBJECT PERCEPTION AND MEMORY
                Empirical evidence
Attention capture:

Recently, Schreij, Owens, and Theeuwes (2008) reported that
abruptly occurring distracters produce costs in performance even in
the presence of a top-down set for colour.

They argue that these results show that abrupt onset of new
objects captures attention independent of a top-down set and thus,
provides conclusive evidence against the idea that attentional
capture is contingent on top-down attentional control settings.




                      HUL 211 OBJECT PERCEPTION AND MEMORY
                Empirical evidence
Scene gist:

Gist of a scene can be gathered even when we see 7 images per
second in an unpredictable random sequence (Potter, 1971; 1975)

This is not even time enough for an eye movement, hence no top
down expectancies can work here.

Adherents of bottom up processing argue that scene gist can be
extracted only on the basis of feed forward processing which yields
a coarse image representation.

Argue against???

                      HUL 211 OBJECT PERCEPTION AND MEMORY
         Thank you




HUL 211 OBJECT PERCEPTION AND MEMORY

				
DOCUMENT INFO
Shared By:
Tags:
Stats:
views:19
posted:5/7/2012
language:English
pages:11
Description: Object Perception and Memory Lecture Series