<p>Initially we pursue a minimal model of a cognitive system. This in turn form the basis for the development of amethodological and theoretical framework. Two methodological requirements of the model are that explanation be from the perspective of the phenomena, and that we have structural determination. The minimal model is derived from the explanatory side of a biologically based cognitive science. Fransisco Varela is our principal source for this part. The model defines the relationship between a formally defined autonomous system and an environment, in such a way as to generate the world of the system, its actual environment. The minimal model is a modular explanation in that we find it on different levels in bio-cognitive systems, from the cell to small social groups. For the latter and for the role played by artefactual systems we bring in Edwin Hutchins' observational study of a cognitive system in action. This necessitates the introduction of a complementary form of explanation. A key aspect of Hutchins' findings is the social domain as environment for humans. Aspects of human cognitive abilities usually attributed to the person are more properly attributed to the social system, including artefactual systems.</p><p>Developing the methodological and theoretical framework means making a transition from the bio-cognitive to the computational. The two complementary forms of explanation are important for the ability to develop a methodology that supports the construction of actual systems. This has to be able to handle the transition from external determination of a system in design to internal determination (autonomy) in operation.</p><p>Once developed, the combined framework is evaluated in an application area. This is done by comparing the standard conception of the Semantic Web with how this notion looks from the perspective of the framework. This includes the development of the methodological framework as a metalevel external knowledge representation. A key difference between the two approaches is the directness by which the semantic is approached. Our perspective puts the focus on interaction and the structural regularities this engenders in the external representation. Regularities which in turn form the basis for machine processing. In this regard we see the relationship between representation and inference as analogous to the relationship between environment and system. Accordingly we have the social domain as environment for artefactual agents. For human level cognitive abilities the social domain as environment is important. We argue that a reasonable shortcut to systems we can relate to, about that very domain, is for artefactual agents to have an external representation of the social domain as environment.</p>

This paper introduces the delta tree, a data structure that represents an object using a set of reference images. It also describes an algorithm for generating arbitrary re-projections of an object by traversing its delta tree. Delta trees are an efficient representation in terms of both storage and rendering performance. Each node of a delta tree stores an image taken from a point on a sampling sphere that encloses the object. Each image is compressed by discarding pixels that can be reconstructed by warping its ancestor's images to the node's viewpoint. The partial image stored at each node is divided into blocks and represented in the frequency domain. The rendering process generates an image at an arbitrary viewpoint by traversing the delta tree from a root node to one or more of its leaves. A subdivision algorithm selects only the required blocks from the nodes along the path. For each block, only the frequency components necessary to reconstruct the final image at an appropriate sampling density are used. This frequency selection mechanism handles both antialiasing and level-of-detail within a single framework. A complex scene is initially rendered by compositing images generated by traversing the delta trees of its components. Once the reference views of a scene are rendered once in this manner, the entire scene can be reprojected to an arbitrary viewpoint by traversing its own delta tree. Our approach is limited to generating views of an object from outside the object's convex hull. In practice we work around this problem by subdividing objects to render views from within the convex hull.

In this paper, we present a static image compression algorithm for very low bit rate applications. The algorithm reduces spatial redundancy present in images by extracting and encoding edge and mean information. Since the human visual system is highly sensitive to edges, an edge-based compression scheme can produce intelligible images at high compression ratios. We present good quality results for facial as well as textured, 256~x~256 color images at 0.1 to 0.3 bpp. The algorithm described in this paper was designed for high performance, keeping hardware implementation issues in mind. In the next phase of the project, which is currently underway, this algorithm will be implemented in hardware, and new edge-based color image sequence compression algorithms will be developed to achieve compression ratios of over 100, i.e., less than 0.12 bpp from 12 bpp. Potential applications include low power, portable video telephones.

In this paper we describe a new efficient algorithm for recognizing 3D objects by combining photometric and geometric invariants. Some photometric properties are derived, that are invariant to the changes of illumination and to relative object motion with respect to the camera and/or the lighting source in 3D space. We argue that conventional color constancy algorithms can not be used in the recognition of 3D objects. Further we show recognition does not require a full constancy of colors, rather, it only needs something that remains unchanged under the varying light conditions sand poses of the objects. Combining the derived color invariants and the spatial constraints on the object surfaces, we identify corresponding positions in the model and the data space coordinates, using centroid invariance of corresponding groups of feature positions. Tests are given to show the stability and efficiency of our approach to 3D object recognition.

Novel algorithms for object recognition are described that directly recover the transformations relating the image to its model. Unlike methods fitting the typical conventional framework, these new methods do not require exhaustive search for each feature correspondence in order to solve for the transformation. Yet they allow simultaneous object identification and recovery of the transformation. Given hypothesized % potentially corresponding regions in the model and data (2D views) --- which are from planar surfaces of the 3D objects --- these methods allow direct compututation of the parameters of the transformation by which the data may be generated from the model. We propose two algorithms: one based on invariants derived from no higher than second and third order moments of the image, the other via a combination of the affine properties of geometrical and the differential attributes of the image. Empirical results on natural images demonstrate the effectiveness of the proposed algorithms. A sensitivity analysis of the algorithm is presented. We demonstrate in particular that the differential method is quite stable against perturbations --- although not without some error --- when compared with conventional methods. We also demonstrate mathematically that even a single point correspondence suffices, theoretically at least, to recover affine parameters via the differential method.

We introduce a new method to describe, in a single image, changes in shape over time. We acquire both range and image information with a stationary stereo camera. From the pictures taken, we display a composite image consisting of the image data from the surface closest to the camera at every pixel. This reveals the 3-d relationships over time by easy-to-interpret occlusion relationships in the composite image. We call the composite a shape-time photograph. Small errors in depth measurements cause artifacts in the shape-time images. We correct most of these using a Markov network to estimate the most probable front surface, taking into account the depth measurements, their uncertainties, and layer continuity assumptions.

A common objective in learning a model from data is to recover its network structure, while the model parameters are of minor interest. For example, we may wish to recover regulatory networks from high-throughput data sources. In this paper we examine how Bayesian regularization using a Dirichlet prior over the model parameters affects the learned model structure in a domain with discrete variables. Surprisingly, a weak prior in the sense of smaller equivalent sample size leads to a strong regularization of the model structure (sparse graph) given a sufficiently large data set. In particular, the empty graph is obtained in the limit of a vanishing strength of prior belief. This is diametrically opposite to what one may expect in this limit, namely the complete graph from an (unregularized) maximum likelihood estimate. Since the prior affects the parameters as expected, the prior strength balances a "trade-off" between regularizing the parameters or the structure of the model. We demonstrate the benefits of optimizing this trade-off in the sense of predictive accuracy.

We study the frequent problem of approximating a target matrix with a matrix of lower rank. We provide a simple and efficient (EM) algorithm for solving {\\em weighted} low rank approximation problems, which, unlike simple matrix factorization problems, do not admit a closed form solution in general. We analyze, in addition, the nature of locally optimal solutions that arise in this context, demonstrate the utility of accommodating the weights in reconstructing the underlying low rank representation, and extend the formulation to non-Gaussian noise models such as classification (collaborative filtering).

In this paper, we present an approach to discretizing multivariate continuous data while learning the structure of a graphical model. We derive the joint scoring function from the principle of predictive accuracy, which inherently ensures the optimal trade-off between goodness of fit and model complexity (including the number of discretization levels). Using the so-called finest grid implied by the data, our scoring function depends only on the number of data points in the various discretization levels. Not only can it be computed efficiently, but it is also independent of the metric used in the continuous space. Our experiments with gene expression data show that discretization plays a crucial role regarding the resulting network structure.

The dream of pervasive computing is slowly becoming a reality. A number of projects around the world are constantly contributing ideas and solutions that are bound to change the way we interact with our environments and with one another. An essential component of the future is a software infrastructure that is capable of supporting interactions on scales ranging from a single physical space to intercontinental collaborations. Such infrastructure must help applications adapt to very diverse environments and must protect people's privacy and respect their personal preferences. In this paper we indicate a number of limitations present in the software infrastructures proposed so far (including our previous work). We then describe the framework for building an infrastructure that satisfies the abovementioned criteria. This framework hinges on the concepts of delegation, arbitration and high-level service discovery. Components of our own implementation of such an infrastructure are presented.