George Mason University & The University Of Pittsburgh: Anthony Harrison & Christian Schunn
This document was uploaded by user and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this DMCA report form. Report DMCA
Overview
Download & View George Mason University & The University Of Pittsburgh: Anthony Harrison & Christian Schunn as PDF for free.
More details
- Words: 1,258
- Pages: 1
p e r n a c t io
w e n nd e t t a > (P =goal l a o g a is ep r c> p r o l p l a w ne is u v d = c n o e l t t l a a u > (P is a v is e w =goal on l n i a t s c o u a g t a on st is a i t a t n e se r p e r c> o l l a =visu al-loc u ==> isa vis ew > n n o s i t u c t a p + sta e r r c a n o e o l i n t l p c a h r a t is u np sa v i dep o i = t c a t - lo en l s a e n r i t p re re
p e r n o ti c a w -ne d n e t (P at =goal> l a o g a is ep r c> p o r l p l a w is u ne v d = n oc e l t l t a a u > (P isa vis ew =goal n o n l i t a s c o u a t g n o st a is a i t a t se n e r p e r > c o l al =visu al-loc u ==> is a v is e w n> n o s i t u c t a a + st ep r r c a n o e o l i n t l p c a h r a t u is np v isa dep o i = t c a t o l -l en s a e n r i t p e r re
ACT-R/S
A neurologically inspired model of spatial reasoning
) ==> +prp> rep c o l p l r a isa p loc =visu lr e t in a
Anthony Harrison & Christian Schunn
) > = = +prp> rep lo c p l r a p u s vi isa = c o l la n i t e r
George Mason University & The University of Pittsburgh
)
)
The spatial model is being inspired by numerous areas of spatial research: neurological, developmental, as well as traditional cognitive psychology. It is hoped that these converging lines of evidence will inform a more plausible and defensible model of spatial reasoning and processing. Data and theories from these various domains are presented briefly with respect to their impact on the model. The first structural decision for this model was the separation of visio-spatial representations into three distinct spatial representations and one visual. These divisions follow Previc’s neurological model of 3-D space. The visual representation system (focal space in Previc) already exists in ACT-R/PM, and will be modified only slightly. The remaining spatial representations are Previc’s ambient extrapersonal, action extrapersonal, and peripersonal spaces. Each of these four representational systems will have their own buffers with which to operate. This will allow the model to process incoming representations in parallel (much like the motor and basic visual system operate in parallel in ACT-R/PM).
Peripersonal Space • Region that is immediately within grasp of the individual. • Spans central 60 degrees on the visual field • Lower field bias • Extends to the maximum arm’s reach. • Neuroimaging studies show used during fine motor control and object manipulation • Neuroimaging studies show used during mental rotation • suggests region contains very precise three-dimensional metric representations of objects to be manipulated.
Coordinate System
• Only region with a variable extent and a non-anchored span within the visual field • The majority of cognitive psychology research into vision has focused on this region • It is responsible for object identification, tracking, and reading. • Utilizes much of Byrne’s ACT-R/PM extension.
Encoding Two objects I’ve manipulated: 1. The chair 2. My umbrella
Chair-back isa prp-rep geon chair-back-cube
Coordinate System • Origin • Upper-left of scene (ACT-R/PM) • Object • Feature based composite • Location • Screen-X, Screen-Y • Units •
Umbrella isa prp-rep geon umbrella-cylinder
Chair-seat isa prp-rep geon chair-seat-cube
Encoding 1. 2.
• An episodic memory system is required for most basic spatial reasoning processing. A symbolic mechanism will be used. • Spatial representations are all with respect to an object and the viewer • Spatial relationships allow the tying together of multiple representations within a region (focal to focal, ambient to ambient) • Each region has its own buffer which capacity is limited by activation levels
• Used as general qualitative, locational encoding. • Used to integrate representation across large expanses
Action Space •Used for navigational representations and scene memory. •We borrowed from robotic navigation to determine the functional necessities. • Neuropsychology brain damage studies suggest a representation that uses landmark’s position relative to the observer’s point of gaze.
• Origin • Center of region • Object • Rough 3D geon of spatial extent • Location (of self) • Sector based • Units • Categorical
• Navigation is performed by trying to match up relative positions to landmarks, while avoiding other obscuring landmarks.
Encoding Two distinct geometric regions: 1. My office 2. The hall
Coordinate System • Origin • Center of gaze • Object • Rectangle obstacles • Location • Vectors to obstacle edges • Units • Angular vectors
My-office isa ambient-rep location center geon box-01 Office-hall isa ambient-rep location nil geon box-02
Key References
Other Considerations • Object representations and object identities are separate entities. However, once the identity is established it is linked to the representation as long as it remains active
• Used in postural control (effectively determining which way is up)
Coordinate System
Focal Space
• Origin • Upper torso • Object • Accurate 3D geon • Location (of object) • Vector from upper torso • Units • Psychological cubits
Ambient Space
No, I have no new email Oh, that guy is Lelyn Lelyn isa visual-object location loc-02 …
Screen isa visual-object location loc-04 ….
Cabinet isa action-rep retinal-loc vis-loc1 right-theta 350o right-distance 3m left-theta 300o left-distance 3.1m region My-office
Chair isa action-rep retinal-loc vis-loc2 right-theta 5o right-distance 1m left-theta 355o left-distance 0.5m top-theta 100o top-distance 1.05m region My-office
Encoding Objects that I need to navigate around: 1. A chair 2. A table 3. A cabinet 4. Some guy
Table isa action-rep retinal-loc vis-loc3 right-theta 120o right-distance 0.2m left-theta 10o left-distance 2.5m top-theta 105o top-distance 0.2m region My-office
• Anderson, J. R., & Lebiere, C. (1998). Atomic components of thought. Mahwah, NJ: Erlbaum. • Biederman, I. (1987). Recognition-by-components: A theory of human image understanding. Psychological Review, 94(2), 115117. • Huttenlocher, J., Newcombe, N., & Sandberg, E. H. (1994). The coding of spatial location in young children. Cognitive Psychology, 27(2), 115-148. • Margules, J., & Gallistel, C. R. (1988). Heading in the rat: Determination by environmental shape. Animal Learning & Behavior, 16(4), 404-410. • Presson, C. C. (1982). Strategies in spatial reasoning. Journal of Experimental Psychology: Learning, Memory, & Cognition, 8(3), 243-251. • Previc, F. H. (1998). The neuropsychology of 3-D space. Psychological Bulletin, 124(2), 123-164. • Sandberg, E. H., Huttenlocher, J., & Newcombe, N. (1996). The development of hierarchical representation of two-dimensional space. Child Development, 67, 721-739. • Tourekzky, D. S., & Redish, A. D. (1996). Theory of rodent navigation based on interacting representations of space. Hippocampus, 6, 247-270.
w e n nd e t t a > (P =goal l a o g a is ep r c> p r o l p l a w ne is u v d = c n o e l t t l a a u > (P is a v is e w =goal on l n i a t s c o u a g t a on st is a i t a t n e se r p e r c> o l l a =visu al-loc u ==> isa vis ew > n n o s i t u c t a p + sta e r r c a n o e o l i n t l p c a h r a t is u np sa v i dep o i = t c a t - lo en l s a e n r i t p re re
p e r n o ti c a w -ne d n e t (P at =goal> l a o g a is ep r c> p o r l p l a w is u ne v d = n oc e l t l t a a u > (P isa vis ew =goal n o n l i t a s c o u a t g n o st a is a i t a t se n e r p e r > c o l al =visu al-loc u ==> is a v is e w n> n o s i t u c t a a + st ep r r c a n o e o l i n t l p c a h r a t u is np v isa dep o i = t c a t o l -l en s a e n r i t p e r re
ACT-R/S
A neurologically inspired model of spatial reasoning
) ==> +prp> rep c o l p l r a isa p loc =visu lr e t in a
Anthony Harrison & Christian Schunn
) > = = +prp> rep lo c p l r a p u s vi isa = c o l la n i t e r
George Mason University & The University of Pittsburgh
)
)
The spatial model is being inspired by numerous areas of spatial research: neurological, developmental, as well as traditional cognitive psychology. It is hoped that these converging lines of evidence will inform a more plausible and defensible model of spatial reasoning and processing. Data and theories from these various domains are presented briefly with respect to their impact on the model. The first structural decision for this model was the separation of visio-spatial representations into three distinct spatial representations and one visual. These divisions follow Previc’s neurological model of 3-D space. The visual representation system (focal space in Previc) already exists in ACT-R/PM, and will be modified only slightly. The remaining spatial representations are Previc’s ambient extrapersonal, action extrapersonal, and peripersonal spaces. Each of these four representational systems will have their own buffers with which to operate. This will allow the model to process incoming representations in parallel (much like the motor and basic visual system operate in parallel in ACT-R/PM).
Peripersonal Space • Region that is immediately within grasp of the individual. • Spans central 60 degrees on the visual field • Lower field bias • Extends to the maximum arm’s reach. • Neuroimaging studies show used during fine motor control and object manipulation • Neuroimaging studies show used during mental rotation • suggests region contains very precise three-dimensional metric representations of objects to be manipulated.
Coordinate System
• Only region with a variable extent and a non-anchored span within the visual field • The majority of cognitive psychology research into vision has focused on this region • It is responsible for object identification, tracking, and reading. • Utilizes much of Byrne’s ACT-R/PM extension.
Encoding Two objects I’ve manipulated: 1. The chair 2. My umbrella
Chair-back isa prp-rep geon chair-back-cube
Coordinate System • Origin • Upper-left of scene (ACT-R/PM) • Object • Feature based composite • Location • Screen-X, Screen-Y • Units •
Umbrella isa prp-rep geon umbrella-cylinder
Chair-seat isa prp-rep geon chair-seat-cube
Encoding 1. 2.
• An episodic memory system is required for most basic spatial reasoning processing. A symbolic mechanism will be used. • Spatial representations are all with respect to an object and the viewer • Spatial relationships allow the tying together of multiple representations within a region (focal to focal, ambient to ambient) • Each region has its own buffer which capacity is limited by activation levels
• Used as general qualitative, locational encoding. • Used to integrate representation across large expanses
Action Space •Used for navigational representations and scene memory. •We borrowed from robotic navigation to determine the functional necessities. • Neuropsychology brain damage studies suggest a representation that uses landmark’s position relative to the observer’s point of gaze.
• Origin • Center of region • Object • Rough 3D geon of spatial extent • Location (of self) • Sector based • Units • Categorical
• Navigation is performed by trying to match up relative positions to landmarks, while avoiding other obscuring landmarks.
Encoding Two distinct geometric regions: 1. My office 2. The hall
Coordinate System • Origin • Center of gaze • Object • Rectangle obstacles • Location • Vectors to obstacle edges • Units • Angular vectors
My-office isa ambient-rep location center geon box-01 Office-hall isa ambient-rep location nil geon box-02
Key References
Other Considerations • Object representations and object identities are separate entities. However, once the identity is established it is linked to the representation as long as it remains active
• Used in postural control (effectively determining which way is up)
Coordinate System
Focal Space
• Origin • Upper torso • Object • Accurate 3D geon • Location (of object) • Vector from upper torso • Units • Psychological cubits
Ambient Space
No, I have no new email Oh, that guy is Lelyn Lelyn isa visual-object location loc-02 …
Screen isa visual-object location loc-04 ….
Cabinet isa action-rep retinal-loc vis-loc1 right-theta 350o right-distance 3m left-theta 300o left-distance 3.1m region My-office
Chair isa action-rep retinal-loc vis-loc2 right-theta 5o right-distance 1m left-theta 355o left-distance 0.5m top-theta 100o top-distance 1.05m region My-office
Encoding Objects that I need to navigate around: 1. A chair 2. A table 3. A cabinet 4. Some guy
Table isa action-rep retinal-loc vis-loc3 right-theta 120o right-distance 0.2m left-theta 10o left-distance 2.5m top-theta 105o top-distance 0.2m region My-office
• Anderson, J. R., & Lebiere, C. (1998). Atomic components of thought. Mahwah, NJ: Erlbaum. • Biederman, I. (1987). Recognition-by-components: A theory of human image understanding. Psychological Review, 94(2), 115117. • Huttenlocher, J., Newcombe, N., & Sandberg, E. H. (1994). The coding of spatial location in young children. Cognitive Psychology, 27(2), 115-148. • Margules, J., & Gallistel, C. R. (1988). Heading in the rat: Determination by environmental shape. Animal Learning & Behavior, 16(4), 404-410. • Presson, C. C. (1982). Strategies in spatial reasoning. Journal of Experimental Psychology: Learning, Memory, & Cognition, 8(3), 243-251. • Previc, F. H. (1998). The neuropsychology of 3-D space. Psychological Bulletin, 124(2), 123-164. • Sandberg, E. H., Huttenlocher, J., & Newcombe, N. (1996). The development of hierarchical representation of two-dimensional space. Child Development, 67, 721-739. • Tourekzky, D. S., & Redish, A. D. (1996). Theory of rodent navigation based on interacting representations of space. Hippocampus, 6, 247-270.
Related Documents
University Of Pittsburgh
June 2020 6
George Harrison
May 2020 9
University
June 2020 19
University Of The Punjab
August 2019 35