Catálogo de publicaciones - libros

This paper introduces a novel approach for vision-based head motion tracking and facial expression cloning to create the realistic facial animation of 3D avatar in real time. The exact head pose estimation and facial expression tracking are critical problems to be solved in developing a vision based computer animation. The proposed method consists of dynamic head pose estimation and facial expression cloning. The proposed head pose estimation technique can robustly estimate 3D head pose from a sequence of input video images. Given an initial reference template of head image and corresponding 3D head pose, full the head motion is recovered by projecting a cylindrical head model to the face image. By updating the template dynamically, it is possible to recover head pose robustly regardless of light variation and self-occlusion. In addition, to produce a realistic 3D face animation, the variation of major facial feature points is tracked by use of optical flow and retargeted to the 3D avatar. We exploit Gaussian RBF to deform the local region of 3D face model around the major feature points. During the model deformation, the clusters of the regional feature points around the major facial features are estimated and the positions of the clusters are changed according to the variation of the major feature points. From the experiments, we can prove that the proposed vision-based animation technique efficiently estimate 3D head pose and produce realistic 3D facial animation rather than using feature-based tracking method.

This paper presents a new constrained motion synthesis algorithm for crowd. The algorithm is based on a novel continuous motion graph where infinite number of motion clips can be created by using path editing technique. Because we can create motions with arbitrary trajectories, we can speed up the motion synthesizing time significantly as well as satisfying constraints exactly. The resulting artifact of the motion such as foot skating is solved by the foot skating solver. The algorithm is tested on several different scenarios.

Inspired by ideas from research on geometric and motion levels of detail, we generalize LoD to combine all currently novel techniques on real-time human locomotion generation. Learn from fruitful research on key-framed kinematic methods, physically-based approaches, motion capture data reuse and action synthesis of reactive articulated characters, we design and implement a tentative but viable LoD scheduler. We also developed a data-driven on-line locomotion generation system base on freely-available motion library by integrating this LoD transition scheduler into popular graphics and dynamics game engine. This paper will then give a brief overview on our experimental results and discussion on future work.

Research in virtual characters and techniques that involve them, started several years ago. Nowadays, they can be used in environments which require a high level of interactivity, such as edutainment. A good example of this scenario is the television. This paper presents the application of several techniques related to corporal animation and natural voice recognition in order to achieve a whole body real time virtual presenter for television. Two main goals have been targeted within the application: a high quality of realism of the virtual presenter and real time interaction in a non intrusive way. A first prototype of this work has been tested in a TV program aim at children in the Basque Country television, and currently it is broadcasted daily.

Computer assisted instruction in biomedical engineering (BME) education has been explored and changed dramatically in decades. In this paper we presents a new educational Grid Environment for BME courses. It is designed to be suitable for supporting for the education of medical instrumentation. In emergent situation it helps students to manipulate Rapid Prototyping Machine to fabricate custom artificial human. A rapid prototyping process chain (RPPC), which combines time limits, cost of CT scans, parallelism in computer simulations and bone fabrication, was integrated as workflow in grid service. The grid involves a synthesis of coordinated examinations and large scale, parallel artificial human bones simulations. The portal at students sides need only a thin client and can be accessed by a standard web browser. To illustrate the application of our approach in the grid environment, a femur fabrication case is presented to show its fast and reliable results.

Due to the high-dimensionality of motion captured data which resulted in the complexity in motion analysis, a method of motion data processing based on manifold learning was proposed. Isomap, a classical manifold learning algorithm, was necessary to be improved and extended in this paper. A framework of motion data processing based on manifold learning was built to embed high-dimensionality data into low-dimensionality space. It simplified the motion analysis, and in the same time preserved the original motion features. In order to solve the inefficiency of processing large-scale motion data, Sample Isomap (S-Isomap) algorithm was proposed. Experiments proved that approximate embeddings of motion data computed by S-Isomap were average 10 times faster than by Isomap, while 10% frame samples were selected.

Interactive storytelling in the virtual environment attracts a lot of research interests in recent years. and are two most important elements of a story. Based on these two elements, currently there are two research directions: plot-based and character-based interactive storytelling. However, plot-based approach lacks the refinement of character behaviors as character-based approach. On the other side, character-based approach does not follow a well organized story plot so that the moral of the story might be distorted. Therefore, there is a need to develop an integrated framework to achieve the balance between conveying story moral and enhancing the modeling of character behaviors. In this paper, we propose a hybrid system of the plot-based and character-based approaches. It is constructed as a multi-agent system (MAS), which involves a scriptwriter agent, a director agent, virtual actor agents and other support agents. Fuzzy Cognitive Goal Net (FCGN) is used by the scriptwriter agent to generate story plot, which includes various meaningful storylines. The director agent selects the storyline dynamically and dispatches the scenes to the virtual character agents as behaviors through the decomposing algorithms. With the system, dynamic storylines and character behaviors are generated in realtime based on the audience interactions and context changes. The audiences are able to experience different levels of interactions as an observer, a character or a director. A case study prototype called “mystery illness investigation at nanyang town” has been implemented based on the proposed system.

This study proposes the context model based on multi-agent in learning environment. We show a collaborative model and a scenario among the agents for providing the adaptive learning service in ubiquitous environment. We design the context model based on the learner’s learning states by defining domains for the context model. We also describe the examples for reasoning methods of the context model by using TRIPLE language for reasoning and collaborating between library agent and school agent. We explore possibilities of the context model through the presenting. The proposed context model based on multi-agent will be useful to support the adaptive learning service states for learners.

E-learning has become an important topic in these days both in academia and in industry. Our efforts in this paper concentrate on proposing a conceptual framework to help designer to construct e-learning system. The presented framework can conspicuously help the developer to comprehend the collaboration relationship between the learning objects. Then, the concepts of hierarchical learning process net and hierarchical web services net are defined by hierarchical Petri-net. The model is expressive enough to capture the semantics of learning process and their specificities. By using Petri-net verification technology, the model will have desirable properties. This will definitely help designer to design and optimize the e-learning processes.

The synergistic Web environment provide for collaborative and active learning by its students. To understand the internal beliefs and motivations within a Web-based educational environment, an integrated Web-based learning model was proposed. The model contains the following external variables: courseware, pedagogic design and specification, and the internal beliefs that motivate learning activities. A laboratory experiment was conducted to investigate the influence of coopetitive pedagogic design on learning performance within the integrated Web-based learning model. The valid sample consisted of 138 participants. The results indicate that coopetitive pedagogic design did have an impact on the virtual classroom’s atmosphere (i.e. social presence, perceived competition and cooperation), motivations for learning (i.e. enjoyment and challenge), and learning performance (i.e. flow and performance). The integrated model of Web-based learning is thus an effective model for predicting learning performance