Login Form

Editors

Activity recognition results on UCF Sports and Holywood2

Table above shows the results, obtained on UCF Sports dataset (http://crcv.ucf.edu/data/UCF_Sports_Action.php). We report recognition rate with respect to the number...


Read More...

Computational efficiency and parallel implementation

The developed algorithms are computationally effective and the compositional processing pipeline is well-suited for implementation on massively parallel architectures. Many...


Read More...

Motion hierarchy structure

Our model is comprised of three processing stages, as shown in the Figure. The task of the lowest stage (layers...


Read More...

Server crash

After experiencing a total server failure, we are back online. We apologize for the inconvenience - we are still in...


Read More...

L1: motion features

Layer L1 provides an input to the compositional hierarchy. Motion, obtained in L0 is encoded using a small dictionary.


Read More...
01234

Relevance to the development of science or a scientific field

Successful realization of the research project objectives in scientific terms is directly relevant to the primary scientific field, the field of computer vision, as well as the development of the motion analysis. Learning and motion recognition is one of the central topics of computer, as well as artificial cognitive vision. The reason is the significant scientific challenge and the usability of the solutions for various automated systems based on efficient motion sensing and activity recognition. The planned research is therefore an important step forward in these areas, as it proposes a new paradigm of creating algorithms and systems to detect and interpret motion. The central contribution is a new approach based on learning of hierarchical compositional motion models, which provides a link to a similar concept of perception of the shape category. This leads to a more consistent approach of treating the shape and motion.

In addition to theoretical models we will develop their fast parallel implementations. Our results and methodology will therefore have a potential of being applied in other fields of science, such as the cognitive robotics. In the field of cognitive robotics, the project results will enable faster and more robust detection and interpretation of motion. This will extend the possibility of achieving complex tasks and interactions with the environment and users. Recently, much research in the field of cognitive robotics has focused on the integration of various sources of information into multimodal systems. One of our objectives in the project is to address the shape and motion information within a common theoretical framework. We can therefore expect that our findings can also be used in the subfield of artificial cognitive systems.

Direct impact of the project for the economy and society

Successful realization of the project’s objectives will have a significant impact on the development of information technologies, and consequently the economy and the wider society. Direct relevance to the economy is obvious, as the successful development of new methods for rapid motion analysis will enable the development of many useful products based on observation of an individual’s motion. High­tech companies that build and offer services in the field of video surveillance could use the tools that we will develop within the project to improve their competitiveness. The same technology could be utilized by technologically innovative companies in the field of robotics and artificial cognitive systems, as well as companies that develop advanced human­machine interfaces. Technology, which will be developed within the project, will be ultimately applicable for the implementation of advanced techniques for indexing video databases in the context of motion information (e.g., video collection in media companies). The technology will enable the implementation of affordable smart sensors in telemedicine devices for reporting the unusual behavior. Such sensors, whose central part is an efficient motion analysis, will enable better and more independent life of an aging European population.

Ethical issues

The experimental research involving human subjects (athletes) as part of this proposal will be performed in accordance with the ethical guidelines, laid down by the Ethical Committee of the University of Ljubljana. Participation of the test subjects in all of the studies will be voluntary. Written permissions will be obtained from all participants involved in the study. Furthermore, personal and sensitive data will be treated in accordance with the Slovenian legal act "Zakon o varstvu osebnih podatkov (Uradni list RS, 94/07, uradno prečiščeno besedilo; ZVOP­1)".

Indirect impact of the project for society

Successful realization of the project's objectives will have considerable importance for the promotion of the country, access to foreign skills, participation in international division of labor and education of qualified cadre. Our research team has already developed a specialized system for motion analysis in sports, which is up to date the only non­commercial system based on visual information and is used extensively for research and teaching purposes at the Faculty of Sport University in Ljubljana and the Ruhr­Universität Bochum, Germany. Preliminary work, which is the basis for the proposed project, has already received the recognition at the international scientific meetings and in the form of invited lectures. An example of the research team's general recognition are the review papers from the field of motion analysis in sports (eg, [Barris2008]), who attribute a significant contribution to the development of this field to our research group. Thus it can be reasonably expected that the anticipated results will not only contribute to the promotion of research groups, but will have a broader impact, including the promotion of the country. Education of cadre is multi­layered, ranging from the transfer of scientific advances into the teaching process, to training young researchers and popularization of the scientific achievements in the form of contributions to the institutions for the popularization of science. Examples are the House of Experiments and the Technical Museum of Slovenia, where the group annually contributes in form of the demonstrations of the developed algorithms.

Based on the results we have achieved so far in the modeling, visual learning and inference of the object categories, a spin­off company Suplea Ltd. was founded, to transfer the technology to specific commercial applications. The method has attracted attention in the world and the research team has been invited as a partner in a major U.S. project, led by Teledyne and company­funded agency DARPA (VICOS is the only non­North American partner in the project).

This website uses cookies to manage authentication, navigation, and other functions. By using our website, you agree that we can place these types of cookies on your device.

View e-Privacy Directive Documents

You have declined cookies. This decision can be reversed.

You have allowed cookies to be placed on your computer. This decision can be reversed.