Top-down processing
It is classified as top-down processing when pattern recognition is developed through contextual information. As a result of this approach, one’s perception starts with the most general and gradually progresses to the most specific.
Our perceptions are largely influenced by the expectations we have and prior knowledge. In simple words, based on the knowledge we have, our brain fills in the blanks and anticipates what will come next.
As an example, you are presented with a paragraph that is written in a illegible handwriting style. You will have an easier time understanding what the author wants to say if you read the whole paragraph instead of reading the words separately. The brain may be able to comprehend the gist of a paragraph by using the context provided by the surrounding words.
Gregory’s Theory
According to renowned psychologist Richard Gregory,the perception process is a constructive one that relies on the top-down processing.
Gregory explained that using prior experience and knowledge of a stimulus helps us draw inferences. For him, one’s perception is determined by their best guess or hypothesis regarding the world around them.
When it comes to visual perception, he argues that 90% of the visual information is lost before it reaches the brain. The stimulus is created based on memory and experience that is related to it. Consequently, a perceptual hypothesis is created about it as a result of this event.
In Gregory’s view, incorrect hypotheses can be formed by the brain when faced with a visual illusion, such as the Necker tube, which may result in an inaccurate perception.
HYPOTHESIS AND THEORY article
Front. Psychol., 17 May 2013 | //doi.org/10.3389/fpsyg.2013.00276
What is bottom-up and what is top-down in predictive coding?
Karsten Rauss1* and Gilles Pourtois2
- 1Institute of Medical Psychology and Behavioral Neurobiology, Faculty of Medicine, University of Tübingen, Tübingen, Germany
- 2Psychopathology and Affective Neuroscience Lab, Department of Experimental Clinical & Health Psychology, Ghent University, Ghent, Belgium
Everyone knows what bottom-up is, and how it is different from top-down. At least one is tempted to think so, given that both terms are ubiquitously used, but only rarely defined in the psychology and neuroscience literature. In this review, we highlight the problems and limitations of our current understanding of bottom-up and top-down processes, and we propose a reformulation of this distinction in terms of predictive coding.
Introduction
A central assumption in predictive-coding theories is that activity in the nervous system reflects a process of matching internally generated predictions to external stimulation [Heekeren et al., 2008; Bar, 2009]. Evidence for this assumption has been collected at different levels of neural processing [Rao and Ballard, 1999; Hosoya et al., 2005; Muckli et al., 2005; Summerfield et al., 2006; Alink et al., 2010], which suggests that predictive-coding operates across a wide range of spatial and temporal scales.
A second important assumption in predictive coding is that predictions are transferred from hierarchically higher levels of processing to lower ones, whereas signals traveling in the opposite direction encode prediction errors [Rao and Ballard, 1999; Serences, 2008; Friston, 2009; Grossberg, 2009]. In accordance with terminology commonly used in research on perception, it has thus been suggested that predictive signaling reflects top-down processes, whereas prediction-error signaling constitutes bottom-up processing [Friston, 2009; Alink et al., 2010; Hesselmann et al., 2010].
In this review, we argue that predictive coding provides a powerful conceptual framework that goes beyond the standard dichotomy of “bottom-up” and “top-down.” We first provide an overview of previous attempts at defining bottom-up and top-down processes, and we highlight their problems and limitations. We then outline how predictive coding offers a unique perspective for re-defining what is meant by these terms.
Throughout the remainder of this paper, we refer to ascending and descending connections when discussing the anatomy of biological systems or the architecture of artificial ones [Friston, 2005; Clark, in press]. Functional activity along these two types of connections is referred to as feedforward and feedback, respectively. We avoid additional terms such as “recurrent” or “reentrant,” because we believe they can be subsumed under “feedback” in most cases.
Bottom-Up vs. Top-Down Processing
How exactly do people process perceptual information from the world around them? There are two basic approaches to understanding how this sensation and perception takes place. One of these is known as bottom-up processing and the other is known as top-down processing.
Bottom-Up
Data driven
Focuses on incoming sensory data
Takes place in real time
Top-Down
Info is interpreted using contextual clues
Uses previous experience and expectations
Bottom-up processing can be defined as sensory analysis that begins at the entry-level—with what our senses can detect. This form of processing begins with sensory data and goes up to the brain's integration of this sensory information. Information is carried in one direction starting with the retina and proceeding to the visual cortex.
This process suggests that processing begins with a perception of the stimuli and is fueled by basic mechanisms developed through evolution. Unlike top-down processing, bottom-up processing is purely data-driven and requires no previous knowledge or learning. Bottom-up processing takes place as it happens.
For example, if you see animage of an individual letteron your screen, your eyes transmit the information to your brain, and your brain puts all of this information together.