Dynamic 3D Human Shape Modeling for Intention Prediction from Video Imagery
Small Business Information
303 Bear Hill Road, Waltham, MA, 02451
AbstractInfoscitex proposes an innovative and feasible solution to Dynamic 3D Human Shape Modeling for Intention Prediction from Video Imagery. We will develop a tool to quickly and automatically detect, track, and extract a subject of interest from video imagery. Using our knowledge and experience of anthropometry and biomechanics, we will develop a data-driven dynamic 3D human modeling technology, HM4D (Human Modeling in 4-Dimension). The HM4D will provide a detailed representation of human shape and a realistic description of joint motion. It implements morphing to represent human shape variations and mapping from one pose to another to describe human shape changes due to pose changes and motion. Using HM4D, dynamic 3D models will be constructed from the silhouette sequences that in turn, are derived from markerless video imagery. Combined with model-based motion analysis and supported by normative human shape and motion databases, our technology will be able to identify and predict human behavior and intention. Infoscitex, in collaboration with Dr. James Davis and Dr. Mariofanna Milanova, two experts in computer vision and human modeling, will prove our concept and demonstrate the feasibility of our approaches in Phase I and develop a fully functional dynamic 3D human modeling technology in Phase II. BENEFIT: Our technology will provide a creative solution to the topic and will offer the following benefits: As the core part of technology, the HM4D provides a detailed description of human shape with high anthropometric fidelity and can be readily morphed to represent a wide anthropometric spectrum; The HM4D can be mapped from pose to pose to describe human shape changes, both rigid and non-rigid, resulting from pose change and motion; The HM4D has sound bio-fidelity by incorporating realistic joint constraints. It can be used to analyze human motion and gait; By converting 2D imagery into a 3D dynamic visualization, our technology provides a wealth of information about human shape, pose, gestures, motion, and gait that are unique attributes of behavior and intent. This technology will have very broad applications in both military and commercial sectors. As a key technology for human-borne threat detection; As a screening tool for the security surveillance of public facilities; For war fighters to efficiently identify enemies from a crowd; For human search and rescue; For the entertainment industry to create vivid human figures and animation; For workstation design to improve human-machine interfacing; For the apparel industry to perform virtual design and fitting.
* information listed above is at the time of submission.