Kortenkamp, D. and Weymouth, T. (1994). Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proceedings of the Twelfth National Conference on Artificial Intelligence (AAAI-94).  Home/Search   Document Not in Database   Summary   Related Articles   Check

....can be used to solve tasks other than fuzzy control, let us give a simplified example of a localization algorithm. Such an algorithm, in a more complex form, has been tested in various experiments using a real robot [11] 5 Further uses of coherent functions In the style of current research [12, 13, 14], we perform localization with respect to a topological map, consisting of places connected by paths, with the addition of local metric information, as done for example in [15] In our work, each place is symbolically described by LPL sentences, which relate an abstract description of the place to ....

D. Kortenkamp and T. Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. In 12th National Conference on Artificial Intelligence (AAAI-94), volume 2, pages 979--984, Seattle, Washington, July 31August 4 1994.

....be used to solve tasks other than fuzzy control, let us give a simplified example of a localization algorithm. Such algorithm, in a more complex form, has been tested in various experiments using a real robot [3, 9, 33, 34] 5 FURTHER USES OF COHERENT FUNCTIONS In the style of current research [21, 22, 24], we perform localization with respect to a topological map, consisting of places connected by paths, with the addition of local metric information, as done for example in [14] In our work, each place is symbolically described by LPL sentences, which relate an abstract description of the place to ....

Kortenkamp, D. and Weymouth, T. (1994). Topological mapping for mobile robots using a combination of sonar and vision sensing. In 12th National Conference on Artificial Intelligence (AAAI-94), volume 2, pages 979-984, Seattle, Washington, July 31-August 4.

....are almost identical especially to the limited sensors available to a robot so the majority of these landmarks do not aid navigation. A slightly better approach is to select landmarks in some way, for example by having an experimenter choose landmarks while looking at images taken by the robot [11], or by using some form of statistical learning, e.g. 17, 3] In this paper we address the problem of learning to select landmarks so that a robot can reliably navigate a route without getting lost. The robot should select landmarks that will be perceivable each time it follows the route ....

David Kortenkamp and Terry Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proceedings of the Twelfth National Conference on Artificial Intelligence (AAAI'9d), pages 979-984, Seattle, Washington, 1994.

....that can be used on line. In the first category is the technique of asking the user to define the landmarks before the robot explores. Humans typically select objects that they find easy to recognise, such as doors, or line segments extracted from camera images recorded as the robot travels [6]. These approaches suffer from the same problem the robot s perceptions of the world are very different to those of a human, and so features that the researcher thinks are distinctive may not be noticeable by the robot. In contrast to this, Thrun [13] addressed the problem of Bayesian ....

D. Kortenkamp, T. Weymouth, Topological mapping for mobile robots using a combination of sonar and vision sensing, in: Proceedings of the 12th National Conference on Artificial Intelligence (AAAI'94), Seattle, WA, 1994, pp. 979--984.

....Therefore, it requires a lot of a priori knowledge such as a quantitative computation model to estimate the geometric features from the robots sensor inputs. On the other hand, from the viewpoint of artificial intelligence, topological map construction methods have been actively studied [5, 4, 3, 8, 11]. A topological map is represented as a graph structure, where the nodes correspond to some characteristic or distinctive places the robot visited, and the arcs correspond to the travel paths or motor behaviors connecting the places. Topological map learning is important for artificial ....

Kortenkamp, D., Weymouth, T.: Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proceedings of the Twelfth National Conference on Artificial Intelligence (1994) 979--984

....also inherently batch, and estimating a map with EM requires multiple passes through the entire data set. The algorithm presented here is incremental. Finally, as pointed out above, there exists a range of algorithms that are variants of the incremental maximum likelihood method described above [26, 46, 47, 56, 65, 89, 91, 92, 93]. Viewed in this light, our maximum likelihood estimator is strikingly similar to a large corpus of existing literature. The key di erence between our work and this literature is the integration of posterior pose estimation, which enables our approach to close cycles. Probably the most related ....

D. Kortenkamp and T. Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proceedings of the Twelfth National Conference on Arti cial Intelligence, pages 979-984, Menlo Park, July 1994. AAAI, AAAI Press/MIT Press.

....also inherently batch, and estimating a map with EM requires multiple passes through the entire data set. The algorithm presented here is incremental. Finally, as pointed out above, there exists a range of algorithms that are variants of the incremental maximum likelihood method described above [16, 32, 33, 41, 50, 68, 69, 70, 71]. Viewed in this light, our maximum likelihood estimator is strikingly similar to a large corpus of existing literature. The key difference between our work and this literature is the integration of posterior pose estimation, which enables our approach to close cycles. Probably the most related ....

D. Kortenkamp and T. Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proceedings of the Twelfth National Conference on Artificial Intelligence, pages 979--984, Menlo Park, July 1994. AAAI, AAAI Press/MIT Press.

....often change their position and can prevent the robot from taking a previously planned trajectory. To acquire and maintain a model of the environment is a major research area in mobile robotics. The most frequently used types of maps are metric and topological maps. Topological models, as used in [14, 17], describe the environment at a coarse resolution. Because of the lack of necessary details, these types of maps are only of limited use for visualizations. Metric maps, on the other hand, describe the environment at a finer level of detail. A popular approach are discrete occupancy grids proposed ....

D. Kortenkamp and T. Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proc. of the Twelfth National Conference on Artificial Intelligence, pages 979--984, 1994.

....from the most recent image to the models assigned to the reference images. 4, 17] apply a neural network to learn the position of the robot given a reference image. One advantage of this approach lies in the interpolation between the different positions from which the reference images were taken. [9] extract vertical lines from camera images and combine this information with data obtained from ultrasound sensors to estimate the position of the robot. 13, 21] consider trajectories in the Eigenspaces of features. A recent work presented in [15] uses scale invariant features to estimate the ....

D. Kortenkamp and T. Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proc. of the National Conference on Artificial Intelligence (AAAI), 1994.

....with the movements carried out by the robot between consecutive measurements. The problem to acquire and maintain a model of the environment has been a major research area in mobile robotics. The most frequently used types of models are metric and topological maps. Topological models, as used in [14,20], describe the environment at a coarse resolution and in a graph like structure. Because such models lack important details such as doors and tables, they are only of limited use for estimating the state of dynamic objects. Metric maps, on the other hand, describe the environment at a finer level ....

D. Kortenkamp, T. Weymouth, Topological mapping for mobile robots using a combination of sonar and vision sensing, in: Proceedings of the 12th National Conference on Artificial Intelligence (AAAI-94), Seattle, WA, 1994, pp. 979--984.

....servo like procedures. Several authors have also considered hybrid maps that combine aspects of both metric and topological representations [Ark90, EM92] Kortenkamp and Weymouth considered the use of multiple sensing modalities to instantiate the nodes of a graph like (topological) representation [KW94]. Other work has also considered the theoretical issues involved in fully covering an unknown graph using topological exploration [DP90] In previous work it has been observed that while topological mapping with ambiguous signatures with absolute certainty is infeasible, the use of a single ....

David Kortenkamp and Terry Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. pages 979--984, 1994.

....array, an approach inspired by the occupancy grids developed for map making [13, 14, 34, 35] Position can also be estimated qualitatively: instead of precise geometric estimates, only topological relationships are produced. This approach has been implemented in several indoor robots [22, 55]. The same division between quantitative and qualitative methods is illuminating when classifying the few works that deal with outdoor position estimation. The work of Levitt et al. 30] proposes a number of qualitative constraints that can be used to generate region based estimates for position ....

D. Kortenkamp and T. Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. XII National Conference on Artificial Intelligence, 2:979--984, 1994.

....proposed for environment representation. One of them consists of using grid cells, which allows a simple environment representation to be built from sensor data. 4,18,20] Another solution consists of using topological maps, in which the environment is represented by means of objects and graphs; [6,7,9,14,16] path planning becomes easier, but learning and sensor integration becomes more difficult. Different neural network approaches have been used to maintain the representation of the environment [8,12,15,17,19,22] but many models have been tested in simulated environments, and do not cope with real ....

Kortenkamp, David; and Weymouth, T. 1994. Topological Mapping for Mobile Robots Using a Combination of Sonar and Vision Sensing. In Proccedings of the Twelfth National Conference on Artificial Intelligence, 979984. Menlo Park, Calif.: AAAI Press.

....[5] provides an excellent overview of the state of the art in localization. Localization plays a key role in various successful mobile robot architectures presented in [14, 25, 30, 44, 45, 50, 55, 57, 70] and various chapters in [40] While some localization approaches, such as those described in [31, 41, 62, 34] localize the robot relative to landmarks in a topological map, our approach localizes the robot in a metric space, just like those methods proposed in [2, 65, 68] Almost all existing approaches address single robot localization only. Moreover, the vast majority of approaches is incapable of ....

....as black rectangles with white dots [3] Of course, modifying the environment is not an option in many application domains. Some of the more advanced approaches use more natural landmarks that do not require modifications of the environment. For example, the approaches of Kortenkamp and Weymouth [41] and Mataric [47] use gateways, doors, walls, and other vertical objects to determine the robot s position. The Helpmate robot uses ceiling lights to position itself [36] Dark bright regions and vertical edges are used in [13, 71] and hallways, openings and doors are used by the approaches ....

D. Kortenkamp and T. Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proc. of the National Conference on Artificial Intelligence (AAAI), 1994.

....of the hierarchy are common to the adaptive behavior system [Tunstel, 2001] and BISMARC. This method is but one of many available [see Pirjanian, 1998 and references therein] 4. Map Making Memory Work in biologically inspired systems for robotic navigation has a rich history [Mataric, 1991; Kortenkamp and Weymouth, 1994; Touretzky, et al. 1994] An overview of other recent studies can be found in [Huntsberger and Rose, 1998] Biological navigation strategies can be characterized as a four level hierarchy based on complexity analysis [Trullier, et al. 4 Huntsberger 1997] The levels are: movement in relation ....

Kortenkamp, D. and Weymouth, T. 1994. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proc. Twelfth National Conference on Artificial Intelligence (AAAI-94).

....may reduce mapping time and improve overall efficiency. 6. Related Work Much research has focused on robot exploration and map learning. Many of the original work and subsequent discoveries in grid based mapping have been done by Moravec (1988) Elfes (1989) and recently Yamauchi (1998) Kortenkamp (1994) performed some of the first experiments in gateway usage and determination. Yamauchi and Langley s (1996) work in place learning motivated our approach, but recently Sebastian Thrun (1998) has developed a system for continuous localization that gave us a need to fill. Our approach is designed to ....

Kortenkamp, D., and Weymouth, T. (1994). Topological mapping for mobile robots using a combination of sonar and vision sensing. Proceedings of the Twelfth National Conference on Artificial Intelligence, 979-984.

....by a grid based maps quickly grows with the size of the mapped environment, since the resolution of a grid must be fine enough to capture every salient detail of the world. Maps and Plans 132 In a topological map, the world is represented as a graph[Kuipers and Levitt, 1988, Mataric, 1990, Kortenkamp and Weymouth, 1994]. Nodes in the graph correspond to distinct locations, situations, or landmarks, and two nodes are connected if there exists a direct path between them. The position of the robot in a topological model is not dependent on a global frame of reference as with grid based approaches, but rather is ....

Kortenkamp, D. and Weymouth, T. (1994). Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proceedings of the Twelfth National Conference on Artificial Intelligence, pages 979--984, Menlo Park.

No context found.

Kortenkamp, D. and Weymouth, T. (1994). Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proceedings of the Twelfth National Conference on Artificial Intelligence (AAAI-94).

No context found.

Kortenkamp, D., and Weymouth, T. 1994. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proceedings of the Twelfth National Conference on Artificial Intelligence (AAAI-94).

No context found.

D. Kortenkamp and T. Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proceedings of the Twelfth National Conference om Artificial Intelligence. AAAI Press/MIT Press, July 1994.

No context found.

Kortenkamp, D. and Weymouth, T. Topological mapping for mobile robots using a combination of sonar and vision sensing. in: Proceedings of the Twelfth National Conference on Artificial Intelligence, AAAI. AAAI Press/MIT Press, Menlo Park, 1994, pp. 979--984.

No context found.

D. Kortenkamp and T. Weymouth. Topological mapping for mobile robots using combination of sonar and vision sensing. In Proc. of the AAAI Conf., pages 979--984, Menlo Park, CA, 1994.

No context found.

D. Kortenkamp and T. Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proceedings of the Twelfth National Conference on Artificial Intelligence, pages 979--984, Menlo Park, July 1994. AAAI, AAAI Press/MIT Press.

No context found.

David Kortenkamp and Terry Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proc. of the Twelfth National ConferenceonArti#cial Intelligence, pages 979#984, 1994.

No context found.

D. Kortenkamp and T. Weymouth. Topological mapping for mobile robots using a combination of sonar and vision sensing. In Proceedings of AAAI, 1994.

First 50 documents  Next 50

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