Speed limits: Orientation and semantic context interactions constrain natural scene discrimination dynamics.

Abstract

The visual system rapidly extracts information about objects from the cluttered natural environment. In 5 experiments, the authors quantified the influence of orientation and semantics on the classification speed of objects in natural scenes, particularly with regard to object-context interactions. Natural scene photographs were presented in an object-discrimination task and pattern masked with various scene-tomask stimulus-onset asynchronies (SOAs). Full psychometric functions and reaction times (RTs) were measured. The authors found that (a) rotating the full scenes increased threshold SOA at intermediate rotation angles but not for inversion; (b) rotating object or context degraded classification performance in a similar manner; (c) semantically congruent contexts had negligible facilitatory effects on object classification compared with meaningless baseline contexts with a matching contrast structure, but incongruent contexts severely degraded performance; (d) any object-context incongruence (orientation or semantic) increased RTs at longer SOAs, indicating dependent processing of object and context; and (e) facilitatory effects of context emerged only when the context shortly preceded the object. The authors conclude that the effects of natural scene context on object classification are primarily inhibitory and discuss possible reasons.

Ultra Rapid Object Categorization: Effects of Level, Animacy and Context

“…This is true even for short exposure durations and speeded response times, indicating parallel processing of object and context information. Other experiments have demonstrated that manipulation of color [59], proportions [41] and orientation [43] in scenes hamper the recognition of a target object. We failed to find impaired object categorization with an inverted background.…”

“…We failed to find impaired object categorization with an inverted background. According to Rieger et al [43], orientation effects are strongest if the context is rotated by 90°. They showed that rotation by 180° has a weaker effect in natural scene discrimination and suggest some “orientation compensation” mechanisms that allow faster processing of inverted scenes than for intermediate rotation angles.…”

“…Another factor may be the physical relationship between an object and its background. That is, object-background proportions such as size, position and orientation may influence object categorization [41]–[43]. It remains an open question whether or not such perceptually (physically) incongruent background information affect rapid object categorization processes and whether such physical manipulations differ from semantically manipulations.…”

Animal Detection in Natural Images: Effects of Color and Image Database

“…Furthermore, context (or global scene) congruence has generally been shown to have a significant effect on classification accuracy and speed [26], [27]. In the ANID database, images of animals were taken in their natural habitats or in zoo environments coming close to the natural habitat of the animals.…”

Human Cognition in Context: On the Biologic, Cognitive and Social Reconsideration of Meaning as Making Sense of Action

“…Processing in the Brain (Jantzen et al 2007;van Berkum et al 2008) Cognition, Perception and Emotion Shaped and Filtered by Cultural Context (Choi et al 1999;Heine et al 1999;Markus and Kitayama 1991;Nisbett et al 2001;Norenzayan et al 2002) Ecological/Bounded Rationality Research program (Camerer and Fehr 2006;Hutchinson and Gigerentzer 2005) Brain Development Affected by Social and Cultural Experience (Eisenberg 1999;Schlaug et al 2005;Broad et al 2006) The Primacy of Action in Context (Costall et al 2004;Tremoulet and Feldman 2006;Wilson 2002) Everyday Cognition Studies (Hazlehurst et al 2007;Margrett and Willis 2006;Rogoff and Lave 1984) Social Neuroscience: Social Effects on Brain (Adolphs 2003;Todorov et al 2006;Olsson and Phelps 2007) Contextual Effects on Cognition (Rieger et al 2008;Hagoort and van Berkum 2007) Situated/Distributed/ Embodied social cognition (Nardi 1996;Niedenthal et al 2005) In each cell, three representative examples of the specific research agenda are detailed mind as an area of research in cognitive sciences with remarkable scientific potential, from multi-level methodological and theoretical perspectives, that manifestly are out of reach of the old limits of orthodox computationalism.…”