Yiming Ye

Sensor planning for object search

This thesis studies the almost-unexplored field of sensor planning for object search. Object search is the task of efficiently searching for a given 3D object in a given 3D environment by an agent equipped with a camera for target detection and, if the environment configuration is not known, a method of calculating depth, like stereo or laser range finder. Sensor planning for object search refers to the task of selecting the sensing parameters so as to bring the target into the field of view of the camera and to make the image of the target easily detected by the available recognition algorithms. In this thesis, the task of sensor planning for object search is formulated as an optimization problem. This problem is proved to be NP-Complete, thus an approximate solution employing a one step look-ahead strategy is proposed. This approximation is equivalent to the optimal solution under certain conditions. The search region is characterized by the probability distribution of the presence of the target. The goal is to find the desired object reliably with minimum effort. The control of the sensing parameters depends on the current state of the search region and the detecting ability of the recognition algorithm. The huge space of possible sensing actions is decomposed into a finite set of actions that must be considered. In order to represent the surrounding environment of the camera and to determine efficiently the sensing parameters over time, a concept called the sensed sphere is proposed, and its construction, using a laser range finder, is derived.

A COMPLEXITY-LEVEL ANALYSIS OF THE SENSOR PLANNING TASK FOR OBJECT SEARCH

Object search is the task of searching for a given 3D object in a given 3D environment by a controllable camera. Sensor planning for object search refers to the task of how to select the sensing parameters of the camera so as to bring the target into the field of view of the camera and to make the image of the target to be easily recognized by the available recognition algorithms. In this paper, we study the task of sensor planning for object search from the theoretical point of view. We formulate the task and point out many of its important properties. We then analyze this task from the complexity level and prove that this task is NP‐Complete.

On the Collaborative Object Search Team: A Formulation

This paper gives a formulation of a collaborative object search team and studies the learning, interaction and organization within this multiagent environment. Each team member is assumed to be a mobile platform equipped with an active camera and recognition algorithms so that images of the environment can be taken and analyzed for the target object. The goal of the team is to find the target within a given time constraint. In order to do this, the agents must interact and collaborate with each other and must learn and modify the various cooperation styles based on the search results.

Understanding emergent web regularities with information foraging agents

In Web Intelligence (WI) applications, it is a common practice to record Web log data. What remains a big obstacle as well as a great challenge in log analysis today is how to characterize the underlying user behavior from the obtained data.In this paper, we will focus on how to interpret strong regularities in Web surfing in terms of user decision-making patterns, and present an information foraging agent-based approach to characterizing user behavior.

Agents-supported adaptive group awareness: smart distance and WWWaware

We study how agents can facilitate and mediate interaction, communication, and cooperation among people. We propose the concepts of a smart distance and an awareness network in a distributed collaborative environment. We illustrate the architecture of an agent-mediated collaborative system - the agent-buddy system that can create a sense of group presence and, at the same time, preserve the privacy of each user. Virtual springs systems are used to model the awareness degrees among team members. Each agent makes decisions by considering multiple factors. The goal of the multi-agent team is to minimize the global awareness frustrations with respect to different kinds of tasks. Empirical studies were conducted to analyze the influence of individual behavior on global performance for various kinds of tasks.

Smart distance and WWWaware: a multi-agent approach

We propose the concept of a smart distance and the concept of an awareness network in a distributed collaborative environment. We illustrate the architecture of an Agent Mediated Collaborative system, the Agent- Buddy system, that can create a sense of group presence and at the same time preserve the privacy of each user. Virtual springs systems are used to model the awareness degrees among team members. Each agent makes decisions by considering multiple factors. The goal of the multiagent team is to minimize the global awareness frustrations with respect to different kinds of tasks. Empirical studies have been conducted to analyze the influence of individual behavior on global performance for various situations.

KNOWLEDGE GRANULARITY SPECTRUM, ACTION PYRAMID, AND THE SCALING PROBLEM

In this paper we introduce the concept of knowledge granularity and study the relationship between different knowledge representation schemes and the scaling problem. By scale to a task, we mean that an agents planning system and knowledge representation scheme are able to generate the range of behaviors required by the task in a timely fashion. Action selection is critical to an agent performing a task in a dynamic, unpredictable environment. Knowledge representation is central to the agents action selection process. It is important to study how an agent should adapt its methods of representation such that its performance can scale to different task requirements. Here we study the following issues. One is the knowledge granularity problem: to what detail should an agent represent a certain kind of knowledge if a single granularity of representation is to be used. Another is the representation scheme problem: to scale to a given task, should an agent represent its knowledge using a single granularity or a set of hierarchical granularities.

Agent-Based Characterization of Web Regularities

In Web Intelligence (WI) applications, it is common practice to record Web log data What remains a great challenge in Web log mining is how to characterize the underlying user behavior from the obtained data In this chapter, we will focus on how to interpret strong regularities in Web surfing in terms of user decision-making patterns, and present an information foraging agent-based approach to characterizing user behavior. The experimental results based on information foraging agents enable us not only to capture the empirical regularities collected from a real-world Web site, but also to effectively unveil the underlying decision-making mechanisms in user surfing as well as how variables in such mechanisms can affect the empirically observed emergent regularities.

Knowledge granularity and action selection

In this paper we introduce the concept of knowledge granularity and study its influence on an agents action selection process. Action selection is critical to an agent performing a task in a dynamic, unpredictable environment. Knowledge representation is central to the agents action selection process. It is important to study what kind of knowledge the agent should represent and the preferred methods of representation. One interesting research issue in this area is the knowledge granularity problem: to what detail should an agent represent a certain kind of knowledge. In other words, how much memory should an agent allocate to represent a certain kind of knowledge. Here, we first study knowledge granularity and its influence on action selection in the context of an object search agent — a robot that searches for a target within an environment. Then we propose a guideline for selecting reasonable knowledge granularity for an agent in general.

Collective perception in massive, open, and heterogeneous multi-agent environment

The web of interconnected intelligent software agents as well as intelligent hardware agents will be seamlessly embedded in everywhere of our lives and constantly sensing and reacting to the environment. The dynamic and heterogeneous interactions among these agents will provide great opportunities for agent-based services. One of the challenging issues in this agent-based service environment is the task of collective perception: how to make sense of complex sensed data at the conceptual level by a group of collaborative agents. This paper proposes a strategy for collective perception when the agents involved may not share the same knowledge representation or ontology. To avoid the syntax, semantics, and ontological complexities in communicating and understanding among agents, the synthesizing agent collects only the analyzed and categorized results from other agents in the form of a natural number or a vector of natural numbers. It then perform collective perception on top of these categorized results. An eigenspace method is proposed to model and perceive events. Experimental results are presented to show the effectiveness of our mechanism.