T. Fuhr, G. Socher, C. Scheering, and G. Sagerer. A three-dimensional spatial model for the interpretation of image data. In P. Olivier and K.-P. Gapp, editors, Representation and Processing of Spatial Expressions, pages 103--118. Lawrence Erlbaum Associates, 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....with vision capabilities and a psuedo natural language interface. 5.3.1 Background Recently there has been a lot of interest in guided autonomous mobile robots. The use of a natural language command interface is being actively investigated by many [ Bandera et al. 1994; L uth et al. 1994; Fuhr et al. 1995 ] The ability to deal with deictic gestures to localize objects has been demonstrated [ Firby et al. 1995] However, no one has yet explored an interface containing locative spatial expressions that can be used for object localization. We claim that a bit of knowledge and instruction in a ....

....set. Declarative specification of a priori costs compared to training in a specific domain should be investigated. 2. Though we have explored multiple reference frames in the SpatialFEVAHR domain, alternative representations that can be investigated are reference frame independent representations [Fuhr et al. 1995]and object centric representations. 3. The constrained search paradigm is currently being applied only for spatial relations between objects. We need to explore its application to other visually relations, for example, if darker than(A,B) and A is uniquely labeled, then that constrains the value ....

Thomas Fuhr, Gudrun Socher, Christian Scheering, and Gerhard Sagerer. A three-dimensional spatial model for the interpretation of image data. In IJCAI Workshop on Representation and Processing of Spatial Expressions, 1995.

....and a psuedo natural language interface. CHAPTER 5. APPLICATIONS 87 5.3.1 Background Recently there has been a lot of interest in guided autonomous mobile robots. The use of a natural language command interface is being actively investigated by many [ Bandera et al. 1994; L uth et al. 1994; Fuhr et al. 1995 ] The ability to deal with deictic gestures to localize objects has been demonstrated [ Firby et al. 1995] However, no one has yet explored an interface containing locative spatial expressions that can be used for object localization. We claim that a bit of knowledge and instruction in a ....

....set. Declarative specification of a priori costs compared to training in a specific domain should be investigated. 2. Though we have explored multiple reference frames in the SpatialFEVAHR domain, alternative representations that can be investigated are reference frame independent representations [Fuhr et al. 1995]and object centric representations. 3. The constrained search paradigm is currently being applied only for spatial relations between objects. We need to explore its application to other visually relations, for example, if darker than(A,B) and A is uniquely labeled, then that constrains the value ....

Thomas Fuhr, Gudrun Socher, Christian Scheering, and Gerhard Sagerer. A three-dimensional spatial model for the interpretation of image data. In IJCAI Workshop on Representation and Processing of Spatial Expressions, 1995.

....is much more complex. The computational model is able to cope with different frames of reference, represents vague and overlapping meanings of projective relations, and considers the influence of the objects shape on the applicability of prepositions. A detailed description can be found in [4]. The rough idea of the model is presented in figure 7. Instead of detailed geometric models of the objects, we use surrounding boxes as abstractions that are collinear to the objects inertia axes (fig. 7b) A finite number of acceptance volumes is associated with each object (fig. 7c) These are ....

T. Fuhr, G. Socher, C. Scheering, and G. Sagerer. A three-dimensional spatial model for the interpretation of image data. In P. Olivier and K.-P. Gapp, editors, Representation and Processing of Spatial Expressions, pages 103--118. Lawrence Erlbaum Associates, 1997.

....are estimated fitting the projection of all 3D object models to 2D image features in the detected object regions [18] This leads to a 3D scene reconstruction and a calibration of the stereo camera using only the objects in the scene. The 3D information is used to compute spatial relations [4] partitioning the 3D space in an object specific way. According to an intrinsic or deictic meaning of the relations derived from speech understanding and the knowledge of the vantage point of the speaker a degree of fulfillment is calculated for each relation. Shape and size are not computed ....

T. Fuhr, G. Socher, C. Scheering, and G. Sagerer. A threedimensional spatial model for the interpretation of image data. In IJCAI-95 Workshop on Representation and Processing of Spatial Expressions, pages 93--102, Montreal, Canada, 1995. 14 th Int. Joint Conf. on Art. Int. (IJCAI-95).

....reconstruction module provides the concept RK OBJEKT with the pose of each object in 3D (Socher et al. 1995a) The recognition and reconstruction modules are described in section 3.3. The geometric descriptions contained in RK OBJEKTs are used to determine a spatial description of the scene (cf. Fuhr et al. 1995 and section 3.3) To instantiate K IO, one or more reference links to object(s) in the scene must be established. If the utterance contains no localizing subexpression, this reference is found in the K OBJEKT instance that best matches the qualitative characterization of the intended object in K ....

....relations having certain degrees of applicability. The degree of fulfillment of a SR for a given IO with respect to the RO and the RF is determined. Vice versa, given a SR, a RO, and a RF, the corresponding acceptance volumes predict locations in 3D space and by projection also in the 2D images (Fuhr et al. 1995). First psycholinguistic experiments show that subjects strongly accept the system output (Vorwerg et al. 1996) More often than spatial relations, unary object relations as color, size, or shape are used to refer to objects and their properties. Color is a dominant feature for objects in verbal ....

Fuhr, T., Socher, G., Scheering, C., & Sagerer, G. (1995). A three-dimensional spatial model for the interpretation of image data. In IJCAI-95 Workshop on Representation and Processing of Spatial Expressions, Montreal, Canada, pp. 93--102. 14 th Int. Joint Conf. on Art. Int. (IJCAI-95).

....et al. 1996) A model based reconstruction module provides 3D OBJECT with the pose of each object in 3D (Socher et al. 1995) The recognition and reconstruction modules are described in section 3. The geometric descriptions of OBJECTs are used to determine a spatial description of the scene (cf. Fuhr et al. 1995 and section 4) To instantiate DSTEP IO a reference link to a scene object must be established. If the user utterance contains no localizing subexpression, this reference is found in the OBJECT instance that best matches the qualitative characterization of the intended object in DSTEP IO (cf. ....

....relations having certain degrees of applicability. The degree of fulfillment of a SR for a given IO with respect to the RO and the RF is determined. Vice versa, given a SR, a RO, and a RF, the corresponding acceptance volumes predict locations in 3D space and by projection also in the 2D images (Fuhr et al. 1995). First psycholinguistic experiments show that subjects strongly accept the system output (Vorwerg et al. 1996) More often than spatial relations, unary object relations as color, size, or shape are used to refer to objects and their properties. Color is a dominant feature for objects in verbal ....

Fuhr, T., Socher, G., Scheering, C., & Sagerer, G. (1995). A three-dimensional spatial model for the interpretation of image data. In IJCAI-95 Workshop on Representation and Processing of Spatial Expressions, Montreal, Canada, pp.

No context found.

Thomas Fuhr, Gudrun Socher, Christian Scheering, and Gerhard Sagerer: 1995. A three-dimensional spatial model for the interpretation of image data. In IJCAI95 Workshop on Representation and Processing of Spatial Expressions,14 th Int.

No context found.

Fuhr, Thomas, Gudrun Socher, Christian Scheering, and Gerhard Sagerer, 1996. A three-dimensional spatial model for the interpretation of image data. In Representation and Processing of Spatial Expressions, eds. P. Olivier and K. Gapp, Lawrence Erlbaum Associates, 1996.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC