CS564 – Lecture 7 Object Recognition and Scene Analysis Reading

《CS564 – Lecture 7 Object Recognition and Scene Analysis Reading》由会员分享，可在线阅读，更多相关《CS564 – Lecture 7 Object Recognition and Scene Analysis Reading（50页珍藏版）》请在文档大全上搜索。

1、Arbib: CS564 - Brain Theory and Artificial Intelligence, USC, Fall 2001. Lecture 7. Object Recognition CS564 Lecture 7. Object Recognitionand Scene AnalysisReading Assignments:TMB2: Sections 2.2, and 5.2“Handout”: Extracts from HBTNN 2e Drafts: Shimon Edelman and Nathan Intrator: Visual Processing o

2、f Object StructureGuy Wallis and Heinrich Blthoff: Object recognition, neurophysiologySimon Thorpe and Michle Fabre-Thorpe: Fast Visual Processing(My thanks to Laurent Itti and Bosco Tjan for permission to use the slides they prepared for lectures on this topic.)Arbib: CS564 - Brain Theory and Artif

3、icial Intelligence, USC, Fall 2001. Lecture 7. Object Recognition Arbib: CS564 - Brain Theory and Artificial Intelligence, USC, Fall 2001. Lecture 7. Object Recognition Bottom-Up Segmentation or Top-Down Control?Arbib: CS564 - Brain Theory and Artificial Intelligence, USC, Fall 2001. Lecture 7. Obje

4、ct Recognition Object RecognitionWhat is Object Recognition?k Segmentation/Figure-Ground Separation: prerequisite or consequence?k Labeling an object The focus of most studiesk Extracting a parametric description as wellObject Recognition versus Scene Analysisk An object may be part of a scene ork I

5、tself be recognized as a “scene”What is Object Recognition for?k As a context for recognizing something else (locating a house by the tree in the garden)k As a target for action (climb that tree)Arbib: CS564 - Brain Theory and Artificial Intelligence, USC, Fall 2001. Lecture 7. Object Recognition Wh

6、at versus How” in HumanVisualCortexParietalCortexInferotemporalCortexHow (dorsal)What (ventral)reach programminggrasp programmingAT: Goodale and MilnerLesion here: Inability to verbalize or pantomime size or orientationDF: Jeannerod et al.Lesion here: Inability to Preshape(except for objects with si

7、ze “in the semantics”Monkey Data:Mishkin and Ungerleider on“What” versus “Where”Arbib: CS564 - Brain Theory and Artificial Intelligence, USC, Fall 2001. Lecture 7. Object Recognition Clinical StudiesStudies with patients with some visual deficits strongly argue that tight interaction between where a

8、nd what/how visual streams are necessary for scene interpretation.Visual agnosia: can see objects, copy drawings of them, etc., but cannot recognize or name them!Dorsal agnosia: cannot recognize objectsif more than two are presented simulta-neously: problem with localizationVentral agnosia: cannot i

9、dentify objects.Arbib: CS564 - Brain Theory and Artificial Intelligence, USC, Fall 2001. Lecture 7. Object Recognition These studies suggestWe bind features of objects into objects (feature binding)We bind objects in space into some arrangement (space binding)We perceive the scene.Feature binding =

10、what/how streamSpace binding = where streamDouble role of spatial relationships:k To relate different portions of an object or scene as a guide to recognitionk Augmented by other “how” parameters, to guide our behavior with respect to the observed scene.Arbib: CS564 - Brain Theory and Artificial Int

11、elligence, USC, Fall 2001. Lecture 7. Object Recognition Inferotemporal PathwaysLater stages of IT (AIT/CIT) connect to the frontal lobe, whereas earlier ones (CIT/PIT) connect to the parietal lobe. This functional distinction may well be important in forming a complete picture of inter-lobe interac