Upper Body Tracking and 3D Gesture Reconstruction Using Agent-Based Architecture

Authored by Chao Peng, Bing Fang, Francis Quek, Yong Cao, Seung In Park, Liguang Xie

Date Published: 2015

DOI: 10.1142/s0219467815500163

Sponsors: United States National Science Foundation (NSF)

Platforms: No platforms listed

Model Documentation: Other Narrative

Model Code URLs: Model code not found

Abstract

In this paper, we present an upper human body tracking system with agent-based architecture. Our agent-based approach departs from process-centric model where the agents are bound to specific processes, and introduces a novel model by which agents are bound to the objects or sub-objects being recognized or tracked. To demonstrate the effectiveness of our system, we use stereo video streams, which are captured by calibrated stereo cameras, as inputs and synthesize human animations which are represented by 3D skeletal motion data. Different from our previous researches, the new system does not require a restricted capture environment with special lighting condition and projected patterns and subjects can wear daily clothes (we do NOT use any markers). With the success from the previous researches, our pre-designed agents are autonomous, self-aware entities that are capable of communicating with other agents to perform tracking within agent coalitions. Each agent with high-level abstracted knowledge seeks `evidence' for its existence from both low-level features (e.g. motion vector fields, color blobs) as well as from its peers (other agents representing body-parts with which it is compatible). The power of the agent-based approach is the flexibility by which domain information may be encoded within each agent to produce an overall tracking solution.

Tags

multiagent systems Model Human motion