In this paper, we present a landmark saliency architecture (LSA) for automatically selecting candidate landmarks. Using a dataset of 21 outdoor stereo images generated by LSA, we show that the terrain spatiogram representation reliably recognizes automatically selected landmarks. The terrain spatiogram results are shown to improve on two purely appearance based approaches: template matching and image histogram matching.

We have previously presented an approach to landmark representation using Terrain Spatiograms – an extension to image spatiograms in which the spatial information relates to the scene rather than the image. Because this representation preserves depth structure, it is possible to identify and filter potential occlusions.

We present an approach to identifying and filtering occlusions using terrain spatiograms, and report experimental results on 20 landmark datasets for varying states of occlusion. We show that occlusion can be detected and filtered, resulting in improved landmark matching scores.

Our approach is linked to a SOAR-based problem-solving and learning framework. In this paper, we present an architecture for the perceptive and world modelling components of ADAPT and report on experimental results using this architecture to predict complex object behaviour. A novel aspect of our approach is a ‘mirror system’ that ensures that the modelled background and foreground objects are synchronized with observations and task-based expectations. This is based on our prior work on comparing real and synthetic images. We show results for a moving object that collides and rebounds from its environment, hence showing that this perception-based problem solving approach has the potential to be used to predict complex object motions.

In this paper, we present an approach to continually synchronize the real and synthetic video by mapping the affine transformation yielded by the real/synthetic image comparison to a new pose for the synthetic camera. We show a series of results for pairs of real and synthetic scenes containing objects including similar and different scenes.

In this paper we propose an extended MMD operation that relaxes the constraint and allows the real and synthetic scenes to differ in some features but not in (selected) other features. Image difference operations that allow a real image and synthetic image generated from an arbitrarily colored graphical model of a scene to be compared. Scenes with the same content show a zero difference. Scenes with varying foreground objects can be controlled to compare the color, size and shape of the foreground.

We experimentally evaluate a cognitive only and reactive only approach to traversing a building filled with various numbers of challenges and compare their performance. The reactive only approach succeeds only 38% of the time, while the cognitive only approach succeeds 100% of the time. While the cognitive only approach produces very impressive behaviour, our results indicate how much better the combination of cognitive and behaviour-based can be.

In this paper, we report on our initial research for the Defense Threat Reduction Agency (DTRA) in understanding what is required from a performance guarantee to give a mission designer the information necessary to understand how well a robot program will perform in a specific environment. We link this to prior work on metrics for robot performance. Using a simple mission scenario, we explore the implications of uncertainty in the four components of the problem: the robot program, and the sensors, actuators and environment with which the program is executed.