Learning Shape Descriptions: Generating and Generalizing Models of Visual Objects

We present the results of an implemented system for learning structural prototypes from grey-scale images. We show how to divide an object into subparts and how to encode the properties of these subparts and the relations between them. We discuss the importance of hierarchy and grouping in representing objects and show how a notion of visual similarities can be embedded in the description language. Finally we exhibit a learning algorithm that forms class models from the descriptions produced and uses these models to recognize new members of the class.

Stimulus Familiarity Determines Recognition Strategy for Novel 3-D Objects

We describe a psychophysical investigation of the effects of object complexity and familiarity on the variation of recognition time and recognition accuracy over different views of novel 3D objects. Our findings indicate that with practice the response times for different views become more uniform and the initially orderly dependency of the response time on the distance to a "good" view disappears. One possible interpretation of our results is in terms of a tradeoff between memory needed for storing specific-view representations of objects and time spent in recognizing the objects.

The Alignment of Objects With Smooth Surfaces: Error Analysis of the Curvature Method

The recognition of objects with smooth bounding surfaces from their contour images is considerably more complicated than that of objects with sharp edges, since in the former case the set of object points that generates the silhouette contours changes from one view to another. The "curvature method", developed by Basri and Ullman [1988], provides a method to approximate the appearance of such objects from different viewpoints. In this paper we analyze the curvature method. We apply the method to ellipsoidal objects and compute analytically the error obtained for different rotations of the objects. The error depends on the exact shape of the ellipsoid (namely, the relative lengths of its axes), and it increases a sthe ellipsoid becomes "deep" (elongated in the Z-direction). We show that the errors are usually small, and that, in general, a small number of models is required to predict the appearance of an ellipsoid from all possible views. Finally, we show experimentally that the curvature method applies as well to objects with hyperbolic surface patches.

Distance Metric Between 3D Models and 3D Images for Recognition and Classification

Similarity measurements between 3D objects and 2D images are useful for the tasks of object recognition and classification. We distinguish between two types of similarity metrics: metrics computed in image-space (image metrics) and metrics computed in transformation-space (transformation metrics). Existing methods typically use image and the nearest view of the object. Example for such a measure is the Euclidean distance between feature points in the image and corresponding points in the nearest view. (Computing this measure is equivalent to solving the exterior orientation calibration problem.) In this paper we introduce a different type of metrics: transformation metrics. These metrics penalize for the deformatoins applied to the object to produce the observed image. We present a transformation metric that optimally penalizes for "affine deformations" under weak-perspective. A closed-form solution, together with the nearest view according to this metric, are derived. The metric is shown to be equivalent to the Euclidean image metric, in the sense that they bound each other from both above and below. For Euclidean image metric we offier a sub-optimal closed-form solution and an iterative scheme to compute the exact solution.

Linear Object Classes and Image Synthesis from a Single Example Image

The need to generate new views of a 3D object from a single real image arises in several fields, including graphics and object recognition. While the traditional approach relies on the use of 3D models, we have recently introduced techniques that are applicable under restricted conditions but simpler. The approach exploits image transformations that are specific to the relevant object class and learnable from example views of other "prototypical" objects of the same class. In this paper, we introduce such a new technique by extending the notion of linear class first proposed by Poggio and Vetter. For linear object classes it is shown that linear transformations can be learned exactly from a basis set of 2D prototypical views. We demonstrate the approach on artificial objects and then show preliminary evidence that the technique can effectively "rotate" high- resolution face images from a single 2D view.

Learning and Example Selection for Object and Pattern Detection

This thesis presents a learning based approach for detecting classes of objects and patterns with variable image appearance but highly predictable image boundaries. It consists of two parts. In part one, we introduce our object and pattern detection approach using a concrete human face detection example. The approach first builds a distribution-based model of the target pattern class in an appropriate feature space to describe the target's variable image appearance. It then learns from examples a similarity measure for matching new patterns against the distribution-based target model. The approach makes few assumptions about the target pattern class and should therefore be fairly general, as long as the target class has predictable image boundaries. Because our object and pattern detection approach is very much learning-based, how well a system eventually performs depends heavily on the quality of training examples it receives. The second part of this thesis looks at how one can select high quality examples for function approximation learning tasks. We propose an {em active learning} formulation for function approximation, and show for three specific approximation function classes, that the active example selection strategy learns its target with fewer data samples than random sampling. We then simplify the original active learning formulation, and show how it leads to a tractable example selection paradigm, suitable for use in many object and pattern detection problems.

Generating Compliant Motion of Objects with an Articulated Hand

The flexibility of the robot is the key to its success as a viable aid to production. Flexibility of a robot can be explained in two directions. The first is to increase the physical generality of the robot such that it can be easily reconfigured to handle a wide variety of tasks. The second direction is to increase the ability of the robot to interact with its environment such that tasks can still be successfully completed in the presence of uncertainties. The use of articulated hands are capable of adapting to a wide variety of grasp shapes, hence reducing the need for special tooling. The availability of low mass, high bandwidth points close to the manipulated object also offers significant improvements I the control of fine motions. This thesis provides a framework for using articulated hands to perform local manipulation of objects. N particular, it addresses the issues in effecting compliant motions of objects in Cartesian space. The Stanford/JPL hand is used as an example to illustrate a number of concepts. The examples provide a unified methodology for controlling articulated hands grasping with point contacts. We also present a high-level hand programming system based on the methodologies developed in this thesis. Compliant motion of grasped objects and dexterous manipulations can be easily described in the LISP-based hand programming language.

Geometry and Photometry in 3D Visual Recognition

The report addresses the problem of visual recognition under two sources of variability: geometric and photometric. The geometric deals with the relation between 3D objects and their views under orthographic and perspective projection. The photometric deals with the relation between 3D matte objects and their images under changing illumination conditions. Taken together, an alignment-based method is presented for recognizing objects viewed from arbitrary viewing positions and illuminated by arbitrary settings of light sources.

Stimulus Simplification and Object Representation: A Modeling Study

Tsunoda et al. (2001) recently studied the nature of object representation in monkey inferotemporal cortex using a combination of optical imaging and extracellular recordings. In particular, they examined IT neuron responses to complex natural objects and "simplified" versions thereof. In that study, in 42% of the cases, optical imaging revealed a decrease in the number of activation patches in IT as stimuli were "simplified". However, in 58% of the cases, "simplification" of the stimuli actually led to the appearance of additional activation patches in IT. Based on these results, the authors propose a scheme in which an object is represented by combinations of active and inactive columns coding for individual features. We examine the patterns of activation caused by the same stimuli as used by Tsunoda et al. in our model of object recognition in cortex (Riesenhuber 99). We find that object-tuned units can show a pattern of appearance and disappearance of features identical to the experiment. Thus, the data of Tsunoda et al. appear to be in quantitative agreement with a simple object-based representation in which an object's identity is coded by its similarities to reference objects. Moreover, the agreement of simulations and experiment suggests that the simplification procedure used by Tsunoda (2001) is not necessarily an accurate method to determine neuronal tuning.

Grouping For Recognition

This paper presents a new method of grouping edges in order to recognize objects. This grouping method succeeds on images of both two- and three- dimensional objects. So that the recognition system can consider first the collections of edges most likely to lead to the correct recognition of objects, we order groups of edges based on the likelihood that a single object produced them. The grouping module estimates this likelihood using the distance that separates edges and their relative orientation. This ordering greatly reduces the amount of computation required to locate objects and improves the system's robustness to error.

MIT SchMUSE: Class-Based Remote Delegation in a Capricious Distributed Environment

MIT SchMUSE (pronounced "shmooz") is a concurrent, distributed, delegation-based object-oriented interactive environment with persistent storage. It is designed to run in a "capricious" network environment, where servers can migrate from site to site and can regularly become unavailable. Our design introduces a new form of unique identifiers called "globally unique tickets" that provide globally unique time/space stamps for objects and classes without being location specific. Object location is achieved by a distributed hierarchical lazy lookup mechanism that we call "realm resolution." We also introduce a novel mechanism called "message deferral" for enhanced reliability in the face of remote delegation. We conclude with a comparison to related work and a projection of future work on MIT SchMUSE.

Some Aspects of Pattern Recognition by Computer

A computer may gather a lot of information from its environment in an optical or graphical manner. A scene, as seen for instance from a TV camera or a picture, can be transformed into a symbolic description of points and lines or surfaces. This thesis describes several programs, written in the language CONVERT, for the analysis of such descriptions in order to recognize, differentiate and identify desired objects or classes of objects in the scene. Examples are given in each case. Although the recognition might be in terms of projections of 2-dim and 3-dim objects, we do not deal with stereoscopic information. One of our programs (Polybrick) identifies parallelepipeds in a scene which may contain partially hidden bodies and non-parallelepipedic objects. The program TD works mainly with 2-dimensional figures, although under certain conditions successfully identifies 3-dim objects. Overlapping objects are identified when they are transparent. A third program, DT, works with 3-dim and 2-dim objects, and does not identify objects which are not completely seen. Important restrictions and suppositions are: (a) the input is assumed perfect (noiseless), and in a symbolic format; (b) no perspective deformation is considered. A portion of this thesis is devoted to the study of models (symbolic representations) of the objects we want to identify; different schemes, some of them already in use, are discussed. Focusing our attention on the more general problem of identification of general objects when they substantially overlap, we propose some schemes for their recognition, and also analyze some problems that are met.

Attentional Selection in Object Recognition

A key problem in object recognition is selection, namely, the problem of identifying regions in an image within which to start the recognition process, ideally by isolating regions that are likely to come from a single object. Such a selection mechanism has been found to be crucial in reducing the combinatorial search involved in the matching stage of object recognition. Even though selection is of help in recognition, it has largely remained unsolved because of the difficulty in isolating regions belonging to objects under complex imaging conditions involving occlusions, changing illumination, and object appearances. This thesis presents a novel approach to the selection problem by proposing a computational model of visual attentional selection as a paradigm for selection in recognition. In particular, it proposes two modes of attentional selection, namely, attracted and pay attention modes as being appropriate for data and model-driven selection in recognition. An implementation of this model has led to new ways of extracting color, texture and line group information in images, and their subsequent use in isolating areas of the scene likely to contain the model object. Among the specific results in this thesis are: a method of specifying color by perceptual color categories for fast color region segmentation and color-based localization of objects, and a result showing that the recognition of texture patterns on model objects is possible under changes in orientation and occlusions without detailed segmentation. The thesis also presents an evaluation of the proposed model by integrating with a 3D from 2D object recognition system and recording the improvement in performance. These results indicate that attentional selection can significantly overcome the computational bottleneck in object recognition, both due to a reduction in the number of features, and due to a reduction in the number of matches during recognition using the information derived during selection. Finally, these studies have revealed a surprising use of selection, namely, in the partial solution of the pose of a 3D object.

Translucent Procedures, Abstraction without Opacity

This report introduces TRANSLUCENT PROCEDURES as a new mechanism for implementing behavioral abstractions. Like an ordinary procedure, a translucent procedure can be invoked, and thus provides an obvious way to capture a BEHAVIOR. Translucent procedures, like ordinary procedures, can be manipulated as first-class objects and combined using functional composition. But unlike ordinary procedures, translucent procedures have structure that can be examined in well-specified non-destructive ways, without invoking the procedure.

Learning Commonsense Categorical Knowledge in a Thread Memory System

If we are to understand how we can build machines capable of broad purpose learning and reasoning, we must first aim to build systems that can represent, acquire, and reason about the kinds of commonsense knowledge that we humans have about the world. This endeavor suggests steps such as identifying the kinds of knowledge people commonly have about the world, constructing suitable knowledge representations, and exploring the mechanisms that people use to make judgments about the everyday world. In this work, I contribute to these goals by proposing an architecture for a system that can learn commonsense knowledge about the properties and behavior of objects in the world. The architecture described here augments previous machine learning systems in four ways: (1) it relies on a seven dimensional notion of context, built from information recently given to the system, to learn and reason about objects' properties; (2) it has multiple methods that it can use to reason about objects, so that when one method fails, it can fall back on others; (3) it illustrates the usefulness of reasoning about objects by thinking about their similarity to other, better known objects, and by inferring properties of objects from the categories that they belong to; and (4) it represents an attempt to build an autonomous learner and reasoner, that sets its own goals for learning about the world and deduces new facts by reflecting on its acquired knowledge. This thesis describes this architecture, as well as a first implementation, that can learn from sentences such as ``A blue bird flew to the tree'' and ``The small bird flew to the cage'' that birds can fly. One of the main contributions of this work lies in suggesting a further set of salient ideas about how we can build broader purpose commonsense artificial learners and reasoners.