Objectives

AIM@SHAPE is aimed at coordinating research on representing, modelling and processing knowledge related to digital shapes, where by shape it is meant any individual object having a visual appearance which exists in some (two-, three- or higher- dimensional) space (e.g., pictures, sketches, images, 3D objects, videos, 4D animations, etc.). Digital shapes are digital representations of either physically existing objects or virtual objects that can be processed by computer applications. Digital shapes occur and are used in many different environments such as: Industrial Design (e.g., CAD models of products, laser-scanned prototypes), Medical Applications (e.g., tomography or X-rays), Entertainment (e.g., computer animations, virtual actors), Personal Environments (e.g., pictures or videos), Geographical Information Systems (e.g., maps or three-dimensional models of terrains), and many more.   The concept of digital shape therefore encompasses all instances of objects which can be represented in a digital context, irrespectively of their format, use, dimension, provided that they have a geometric nature, that is, they are primarily characterised by their form or spatial extent.

What is common to all shapes is that they all have a geometry (the spatial extent of the object), they can be described by structures (object features and part-whole decomposition), they have attributes (colours, textures, names, attached to an object, its parts and/or its features), they have a semantics (meaning, purpose), and they may also have interaction with time (e.g., history, shape morphing, animation, video).

While the technological advances in terms of hardware and software have made available plenty of tools for using and interacting with the geometry of shapes, the interaction with the semantic content of digital shapes is still far from being satisfactory. While we have tools for viewing digital shapes even in much unspecialised web contexts (e.g. browser plug-ins like SVG or VRML for 3D shapes), we miss tools for interacting with the semantics of digital shapes. It is not possible, for example, to search digital shapes by their semantic meaning. This is partly due to the lack of methods for the automatic extraction of the semantic content of digital shapes (semantic annotation) and partly to the evolution of research on shape modelling which had to be highly focused, in the past years, on the geometric aspects of shapes. The shift from a purely geometric to a semantic-aware level of representation of digital shapes is the ultimate scientific objective of AIM@SHAPE. For the next generation of semantic-enabled systems and services, we foresee a generation of shape models in which knowledge/semantics is explicitly represented and, therefore, can be effectively retrieved, shared, exploited, and used to construct new knowledge.

The scientific objective will be achieved by growing a new multi-disciplinary research field, which deeply integrates Computer Graphics and Vision with Knowledge Technologies. The core of the integration will reside in the homogenisation of the approach to modelling shapes and their associated semantics using knowledge formalisation mechanisms, in particular metadata and ontologies which will provide the rules for linking semantics to shape or shape parts. Through a common formalization framework, it will be possible to build a shared conceptualisation of a multi-layered architecture for shape models , where the simple geometry is organized in different levels of increasing abstraction: geometric , structural and semantic layers.
 

Summary of the project

The multimedia world can be classified into one-dimensional media like text and sound, and multi-dimensional media . Among the latter, those that are characterised by a visual appearance in a space of 2, 3, or more dimensions are called shapes . Examples of shapes are pictures, sketches, images, 3D models of solid objects, videos (disregarding the sound track), 4D (=3D+Time) animations, etc. Shapes are expected to take a central role in the Semantic Web in the next years, with high potential impact in several key areas. Consumer PCs are now all equipped with high-performance 3D graphics hardware. Considering that most of these PCs are connected to the Internet, it seems clear that in the near future 3D data will represent a huge amount of traffic and data stored in the Internet. It has been predicted that geometry is poised to become the fourth wave of digital multimedia communication, where the first three waves were sound in the 1970s, images in the '80s, and video in the '90s. So if the principal use of PCs is currently related to 2D image processing and visualization (if one restricts to multimedia content), the next step is to add new dimensions (3D geometry, 4D, i.e. time-dependent, geometry) to this information content and endow it with knowledge (semantics).  

The mission of AIM@SHAPE is to advance research in the direction of semantic-based shape representations and semantic-oriented tools to acquire, build, transmit, and process shapes with their associated knowledge. We foresee a generation of shapes in which knowledge is explicitly represented and, therefore, can be retrieved, processed, shared, and exploited to construct new knowledge.

• - the formalisation of shape knowledge and the definition of shape ontologies in specific contexts;
• - the definition of shape behaviours which formalise the interoperability between shapes;
• - the delineation of methods for knowledge-based design of shapes;
• - the definition of tools for semantics-dependent mapping of shapes.

This scientific goal of AIM@SHAPE is ambitious, and it can be achieved only by establishing the infrastructures necessary for growing a new multidisciplinary research field, where excellence in Computer Graphics and Vision, Computational Mathematics, Geometric Modelling, Computer Science, CAD and Engineering, joins state-of-the-art Knowledge Technologies.

The scientific innovation sought by AIM@SHAPE will be high in terms of the new methodological approach to model digital shapes. Current information systems may handle the geometric representation of digital shapes, but not their semantics (meaning or functionality) in a given context. Scientific innovation will be also high in terms of proposing and developing ontologies for modelling digital shapes and their semantics in key applied sectors. Up to now, the efforts in this direction only touch the visual appearance of 3D objects (geometry) but not the meaning of the shape in a specific sector.

Technological innovation will be high in terms of tools for the automated semantic annotation of digital shapes, as well as tools for accounting for the semantics while digitising, modelling, and sharing shape data. The innovation will be measured against the existence of a coherent and integrated Digital Shape Workbench (DSW) and its use as an e-Science framework of tools and services for modelling, processing and interpreting digital shapes.

The AIM@SHAPE consortium of 14 excellent research institutions in foundational and applied fields of shape modelling will pursue lasting integration both at the foundational level, by initiating a new Theory of Digital Shapes, and at the component level, by developing a Digital Shape Workbench as a common platform for shape models and software tools. Integrating activities will include the design of a common shape ontology and a program for human capital mobility and training. Spreading of excellence activities will include an international forum, an industrial users' group and regular conferences.

EPFL-VRlab's contribution

Activities within AIM@SHAPE are related to the definition of Ontology for Virtual Humans. EPFL is involved in the development of standards for animation and description of shapes (MPEG-4, MPEG-7). EPFL is also developing technologies related to action with smart object. EPFL is the Workpackage 8 leader which in charge of AIM@SHAPE Openness.

Partners

Website of the project Aim@shape

VRlab contact