Michael J. Daily

Augmented-reality visualizations guided by cognition: perceptual heuristics for combining visible and obscured information

One unique feature of mixed and augmented reality (MR/AR) systems is that hidden and occluded objects an be readily visualized. We call this specialized use of MR/AR, obscured information visualization (OIV). In this paper, we describe the beginning of a research program designed to develop such visualizations through the use of principles derived from perceptual psychology and cognitive science. In this paper we surveyed the cognitive science literature as it applies to such visualization tasks, described experimental questions derived from these cognitive principles, and generated general guidelines that can be used in designing future OIV systems (as well improving AR displays more generally). We also report the results from an experiment that utilized a functioning AR-OIV system: we found that in relative depth judgment, subjects reported rendered objects as being in front of real-world objects, except when additional occlusion and motion cues were presented together.

Color image segmentation using Markov random fields

The use of Markov random fields (MRFs) in color image segmentation of natural outdoor scenes is discussed. MRFs provide an elegant means of specifying a local energy function which embodies the expected dependencies of neighboring pixels and includes both the prior and posterior probabilistic distributions. This local neighborhood-based specification of dependencies avoids ad hoc brittle methods using global image knowledge. A brief analysis of ongoing research in color differencing methods is presented, since they are central to the problem of color segmentation. The authors develop and compare the use of three different lattice structures for coupled MRFs with line and color processes based on squares, hexagons, and triangles, and also discusses current efforts in MRF parameter understanding.<<ETX>>

Visualization tools for Free Flight air-traffic management

Free Flight lets pilots modify their routes in real time. It requires new conflict detection, resolution, and visualization decision support tools. We describe a testbed for building and evaluating such tools.

Visualization of conflicts and resolutions in a "Free Flight" scenario

Free Flight will change todays air traffic control system by giving pilots increased flexibility to choose and modify their routes in real time, reducing costs and increasing system capacity. This increased flexibility comes at the price of increased complexity. If Free Flight is to become a reality, future air traffic controllers, pilots, and airline managers will require new conflict detection, resolution and visualization decision support tools. The paper describes a testbed system for building and evaluating such tools, including its current capabilities, lessons we learned and feedback received from expert users. The visualization system provides an overall plan view supplemented with a detailed perspective view, allowing a user to examine highlighted conflicts and select from a list of proposed solutions, as the scenario runs in real time. Future steps needed to improve this system are described.

Fast pattern matching with time-delay neural networks

We present a novel paradigm for pattern matching. Our method provides a means to search a continuous data stream for exact matches with a priori stored data sequences. At heart, we use a neural network with input and output layers and variable connections in between. The input layer has one neuron for each possible character or number in the data stream, and the output layer has one neuron for each stored pattern. The novelty of the network is that the delays of the connections from input to output layer are optimized to match the temporal occurrence of an input character within a stored sequence. Thus, the polychronous activation of input neurons results in activating an output neuron that indicates detection of a stored pattern. For data streams that have a large alphabet, the connectivity in our network is very sparse and the number of computational steps small: in this case, our method outperforms by a factor 2 deterministic finite state machines, which have been the state of the art for pattern matching for more than 30 years.

Modeling threat behaviors in surveillance video metadata for detection using an Analogical Reasoner

Video surveillance systems are generating much more imagery than can be cost-effectively analyzed by human analysts. One approach to the automated analysis of this imagery is to split the problem into video-to-metadata and metadata-to-interpretation tasks. We describe a system for the metadata-to-interpretation task that automatically extracts propositional graphs from the video metadata and generates graphs for use in an analogical reasoning system to detect threat behaviors that occurred in the original video scene.1 2

Implementation of an analogical reasoning system on a parallel recirculating computer architecture

Analogical reasoning systems can be used to increase the efficiency of planning, infer missing structure in test cases, and classify cases across domains. However, analogical reasoning can be computationally expensive. We describe an approach for implementing the SAGE™ analogical reasoning system on a parallel recirculating computer architecture.1 2

Visualization Concepts for Generating Insight from NAS Simulation Data

¶This paper describes our initial steps to develop new visualization concepts that can generate insight and understanding from National Airspace System (NAS) simulation data. The capacity of the United States’ National Airspace System (NAS) must at least double to handle the projected increase in passenger demand by 2025. To address this challenge, new capacity-enhancing concepts are being developed. These concepts are tested and evaluated on a NAS simulation tool called the Airspace Concept Evaluation System (ACES). Concept developers need improved visualization techniques to better understand ACES simulation outputs, and thereby comprehend the strengths, weaknesses and effects of their capacityincreasing concepts. Examining ACES outputs is a nontrivial task. ACES simulates the entire NAS and generates an enormous amount of data. A single simulation run can include over 60,000 flight segments and output tens of gigabytes of data. Traditional approaches for displaying these outputs fall short of what concept developers need. Existing visualization techniques are straightforward geographic plots of aircraft or their related metrics (density, environmental impact, delay, etc.), which are often overwhelming and not illuminating. For example, drawing 10,000 aircraft at their true locations over the continental United States results in a density that is too high for observers to understand the situation and extract useful information. The contribution of this paper is in describing new visualization concepts for this problem domain, where our visualizations do not rely primarily upon plotting data at their true geographic coordinates. We draw from cognitive science principles, perception, ATM characteristics and information visualization techniques to synthesize new approaches for displaying ACES data. Our goal is to enable the user to detect subtle correlations, patterns, trends and relationships that provide insight. We describe our general strategies and approaches, including the results of asking concept developers what specific questions they wanted a visualization tool to answer. Then we present eight new visualization concepts that reveal different aspects of the NAS simulation data, with preliminary implementations of three concepts. Future work includes evaluation on more datasets. A measure of success will be the ability of these new visualization modes to enable users to see subtle but important patterns, trends, correlations, features, and relationships that they could not previously see by any previous means. The goal is to generate insight and allow observers to find important characteristics that they did not even know to look for initially.

Behavior-based control for an eye-head system

Much of the recent interest in active vision has focussed on the development of novel methods for controlling fast pan/tilt camera mounts, called eye-head systems. Simple real-time processing of input images coupled with fast control has enabled interesting system behaviors. This paper describes the ongoing development of behavior-based control methods for a miniature eye-head system. We first describe the eye-head hardware and image processing system. We then define and present approaches for behavior-based control of the eye-head system. Finally, we discuss results from the use of simple behaviors for verging two cameras on moving objects.

Parameter studies for Markov random field models of early vision

Parameter studies for Markov random field (MRF) models of early vision performed using a DAP array processor are discussed. A simple two-dimensional formulation for MRFs with coupled line processes on a rectangular grid is analyzed. Numerous experimental results are presented for real and synthetic intensity images. We empirically analyze the interrelationships between parameters controlling the degree of smoothness in the solution, the discontinuity threshold, and the importance of input data. Trends are identified, especially with regard to methods for automatically setting parameters. In addition, a comparison of results across different resolutions is made.