An example of application of such a system is to animate easily and realistically a virtual performer in a TV set starting from recorded images of a real player. Further applications like virtual reality immersive collaboration environments and real time immersive storytelling will also be investigated. Medical issues are also interesting in the project.

The innovation and challenges of the project rely both in the recognition system to be developed and in both real-time and non real-time applications that such a system will allow.

Principal goal of HUMODAN project is to produce efficiently, realistically and with low cost the virtual animation of an individual by means of processing only a sequence of images recorded from different cameras and avoiding the use of sensors, markers or special suits.

The project addresses following objectives:

• To develop a new system of human motion recognition/reconstruction without using sensors or markers or any special suit.
• To obtain automatically and rapidly (real-time in some applications) a realistic animation of an individual using a sequence of images from recorded from different cameras.
• To use robust biomechanical analysis.
• To develop specific networking applications, interfaces and plug-ins
• To produce an easy-to-use, user-friendly tool
• System will be easier and cheaper to use. Also, and possibly more important, with respect to future new application areas in which the slowness of the process and the necessity of markers are actual main limitations that prevents from using motion capture technology.

In the basis, it will be developed an innovative system for recognition of human motion based on the most modern techniques of image processing, analysis and synthesis. Besides, the system will be enhanced to recognize and analyze other biped and no-biped beings, like for example pet animals, robots, etc. In addition, system will be able to focus only in a part of the body but with high detail, like for example the hands or the face.

To ensure the widest range of applications, the individual recorded will not wear any type of marker or special suit. To this end, biomechanical models will be constructed using a hierarchical and articulated structure in order to establish a correlation between each structural element of the biomechanical model with the analytical characteristics of the images obtained using different views. Innovative shape or part recognition techniques will be applied. The biomechanical model will include a knowledge database to retain high-level information of the motions.

The biomechanical model will also require developing specific kinematics and dynamic models, and analysis and synthesis tools to support firstly the recognition phase and later the reconstruction and animation phase.

To make the system usable it will be also necessary to develop specific applications and plug-ins to integrate the animation into end users tools such as digital TV production software, animation software and virtual environments like a CAVE. To this end, all developed software from the different subsystems will be integrated using adequate interfaces in order to be able to comply with networking applications and real time requirements.

To demonstrate the results of the project, partners will develop following demonstrators within the scope of workpackage 6:

Table 1-1: List of demonstrators of HUMODAN.
No.	Description	Real Time
D1	Immersive and collaborative interactivity in a Virtual Environment	Yes
D2	TV post-production application	No & Yes
D3	Medical application	No
D4	Digital Storytelling scenario, like a network sportive game or joint visit of an exhibition	Yes

Six milestones have been established to assess periodically the progress of the project. Next table summarises them:

Table 1-2: List of milestones.
Milestone	Description	End of Month
M1	Check that the preparation phase of the project has been done correctly, including the definition of requirements for developers' tasks.	5
M2	Check that the system for capture and visualisation has been correctly developed in term of requirements and specifications.	12
M3	Check that the definition of requirements for demonstration phase has been done correctly.	15
M4	Check that system for non real time applications has been correctly developed in term of requirements and specifications.	22
M5	Check that system for real time applications has been correctly developed in term of requirements and specifications.	24
M6	Final review meeting, which will show and prove that all the objectives and requirements have been fulfilled.	30

There are more expected results than expressed in the objectives and requirements. In few years, the system to be developed and its resident technology will allow new applications that today are not available. It is remarkable its future use in education:

• Education on Medicine: surgery training, anatomy, traumatology, etc.
• Medical applications: non-invasive diagnose of neurological damages affecting the locomotor system.
• Virtual laboratories
• Sports

In addition, it will prove to be useful in training the use of tools or processes in dangerous environments. Besides, the system will open new possibilities for advanced and more natural interfaces between human beings and machines.

The HUMODAN project is being submitted to IST Key Action III - Multimedia Content and Tools. In particular, it is addressed to III.5.1: x-Contents futures. The following comments show how the HUMODAN project contributes to the objectives of this Key Action.

"To provide opportunities for high payoff breakthrough research"
The HUMODAN community will expand, as more people will recognize the value of the HUMODAN project for training and education, health, entertainment, and scientific and business collaboration.
The HUMODAN project will be a full solution to composability that will allow the massively parallel implementation of comprehensive virtual worlds, worlds that have not been seeing before. These different worlds will be seamlessly interconnected into a single Internet-wide environment, with portals allowing users to seamlessly travel between these worlds.

"… future developments …"
One of the major targets of the HUMODAN project is to achieve high quality realism in a networked virtual environment.
The HUMODAN project can be used in the future for creating applications in a wide variety of areas (e.g. education, research, health, commerce) supporting collaborative activities into high quality networked virtual immersive environments (e.g. CAVES, etc) allowing users to browse, visualize and to interact with virtual and real objects.
The HUMODAN project can be used in the future as a platform to create virtual communities around Europe. The participants in this virtual community will have a shared sense of space, a shared sense of presence, a shared sense of time, a way to communicate through visualization, and a way to share items in the environment.

"… explore radically new concepts and paradigms …"
The HUMODAN project will allow multiple users to interact in real-time even though these may be located in different European Countries.
The HUMODAN project aims at a sense of realism and an immersive experience by incorporating realistic 3D graphics and real time human animation.
The technologies of the HUMODAN project can be proved an effective tool in Scientific Visualization. The visualization of large data sets over the Internet will open new frontiers in this area, as the scientists will be able to share and show a model easily and effectively in any place around the globe. Scientists need these kinds of systems not only for their own insight, but also to share and evaluate results with their colleagues.

"… address important problems currently impeding progress in the area of multimedia contents and tools, against a 10 year horizon ..."
One of the major problems today in the developing networked virtual environments is the misuse of network capacity. In order to overcome this problem the HUMODAN project will try to provide reasonable solutions to determine the network capacity needed to run the demonstrations. The test beds of the HUMODAN demos will combine the infrastructure of technology innovation (i.e. use of caves) with a high network bandwidth.
A major problem with current networked virtual environments is the lack of high-level security between the users. The HUMODAN project will be seriously involved in providing high-level security to its users. Users will demand protection of personal information that they bring with them and distribute within the HUMODAN.

The project contributes to global aims of Key Action III, which pursues to improve the functionality, usability and acceptability of future information products and service, …, to stimulate creativity, and to enhance education and training systems for lifelong learning.