In this report you find contributions of several partners from the SMASH project. The impact, exploitation and dissemination is very specific for each partner and depends also on the kind of partners. Much effort has been spent to disseminate the results of the project via papers which have been presented at important conferences or have been submitted to journals. A list of all publications can be found in section 9. The separate contributions of the partners explain in sufficient detail their view on the SMASH project and its exploitation possibilities.
The impact of the project has been on several aspects. The most important one has been achieved with the final demonstrator of an integrated system which is controllable with a conventional remote control, using an advanced user interface which brings all the SMASH features to live. This demonstration to several people has resulted in the belief that a SMASH system is both very interesting for the consumer but also that it is realizable with current day technology.
The exploitation of the project will be along different ways. SMASH has developed many new technologies and aspects of a storage system and it will be possible to take some of the developed ideas and try to realize them in a product. In some cases the tape drive will not be necessary and a product may be developed with a hard disk only. On the other end, when developing new tape products, not always a hard disk is needed to implement some SMASH ideas. Several exploitation possibilities exist and will be pursued.
The dissemination of information has been achieved internally in the company by organizing special workshops and many demonstrations. We have realized also a mobile version of SMASH with a mobile satellite dish and we are able to now give demonstrations everywhere. The only infrastructure we need is a power supply plug of 220 volts. With this demonstrator we plan to give some demonstrations at conferences.
The results that have been achieved with the detection, extraction and organization of key frames, especially for movies and broadcast news, have been widely disseminated through contributions to and (invited) presentations at technical conferences in the field of image/video processing and multimedia, and have led to submitted journal papers (see "List of Published Papers"). These results will be exploited by the University partners in the following ways:
The overall impact of the results in detection, extraction and organization of key frames is twofold. Internally they have led to a much better understanding of the possibilities and limitations of using key frames, and have initiated a wealth of new research questions, both technically (e.g., "how to go from key frames to key objects") and cognitively (e.g., "what are representative key frames from a user point of view"). Externally, the concepts of logical story units in movies and topic delineation in container-like programs are novel results that could influence the standardization of "content descriptors" (MPEG-7, DAVIC) in the near future.
The results that have been achieved with watermarking of MPEG compressed video have been widely disseminated through contributions to and (invited) presentations at technical conferences in the field of signal/image/video processing and cryptography, and have led to submitted journal papers (see "List of Published Papers"). These results will be exploited by the University partners in the following ways:
The impact of this work is twofold. In the first place, as a University a better understanding has been obtained of the possibilities and limitations of securing digital video information in general. It has become clear that among the spectrum of potential applications for watermarking, really only copy protection and finger printing of digital video information will have a future. Secondly, we have contributed an entirely new and very efficient watermarking technique for MPEG compressed video streams. The approach set forth in this technique will have an impact on further work in this field by third parties.
Impact:
The structure of the Digital Interface IEEE1394 is constructed within several layers: The specification of the Physical Layer and the specifications and protocols of the Link-, Transaction- and the Bus Management Layer has successfully been verified. It has been shown, that the protocols for the real time transmission of MPEG2 Transport Stream in accordance to the rules described in the international standard IEC61883 are suitable to connect DVB Set Top Boxes with Storage Devices. An Application Programming Interface and the mapping of the commands on the structures of the interface has been specified.
Exploitation & Dissemination:
The knowledge achieved on the digital interface in the course of the project is the basis for future interconnects between different CE instruments as well as PC products. The reference implementation of the interconnect also enabled the company to demonstrate the functional blocks of a full chain for the some key applications
The verification and testing of the Digital Interface will be continued in co-operation with Philips within the Eureka Commend project
It is intended to integrate a Digital Interface in accordance to IEEE1394 into future CE products. The expected time frame for ramp up in the market is in the magnitude of 2 to 3 years.
The project has offered through the WEB-VCR prototyping activity the opportunity to the Olivetti team the opportunity to acquire/deepen the knowledge and master a set of innovative tools dealing with Java, Netscape environment and development tools, MPEG and storing COMBO subsystems integrated in the application prototype. This experience has set up the base for the information gathering in Internet for collecting Internet pages with all its first level links and it has been a first step in the direction of internal development of information gathering agents in an important information domain such as "the news domain". This experience has been exploited internally and will be exploited further in the continuation phase of SMASH project, i.e. STORit project where a service (based on AMICO: http://amico.arianna.it/ ) will be set up for providing to the students and professors an information environment focused on a specific educational domain offering in particular information gathering, filtering and profiling capabilities. In the information gathering and in particular in the preprocessing phase a number of agents, running in background, visit a list of sites selected from an educational domain collecting Internet pages and storing them together with all their first level links in the COMBO subsystems. Later on in the post-processing phase the gathered information are retrieved from the COMBO subsystems for elaboration, indexing and categorization and made available to the user for retrieving according to his own profile.
Tape Storage in Multimedia Applications
Tandberg Data (TD) is traditionally focused on Tape BackUp and Tape archive applications. In SMASH Tandberg has worked together with companies with an entirely different focus and this has been very educating.
The SMASH project has shown that in Multimedia applications the focus is shifted away from the traditional requirements as capacity and transfer rate towards another set of important requirements.
It seems that the most important requirement is reduced Data Access Time. A tape drive can never compete with a hard disk, but it is still necessary to minimise the access time. To reduce the access time one has to work with several functionalities of the tape drive:
Host Interface
Most of today's tape drives supports the SCSI interface. In future Multimedia applications an IEEE1394 interface would be more suitable.
Standardisation
During the SMASH project
Tandberg Data has been an active member of the QIC committee, focusing on the MLR
standards. As the QIC committee now is less active we still have the MLR standards and the
Travan standards that succeeds QIC.
Particularly important for Tandberg Data are the MLR standards (MLR1 and MLR3) since they
have formed a base for future MLR standards such as MLR5 and MLR7.
The results that have been achieved with the remote-education application (REA) have been disseminated through contributions at the technical conferences in the field of multimedia and education, and through local presentations at the University as well as at the Infos '98 exhibition in Slovenia. These results will be exploited by the University and its partners in the following ways:
The impact of the work is manifold. In the first place, we have participated in a large European R&D project with partners from the industry and from universities. The results of such co-operation are new contacts with European partners and important experiences about the work in such projects. Secondly, we have had possibilities to use the latest technology as well as technology which is developing and will be available in the near future. And finally, the remote-education application, which has been developed as a demo application, combines a mass storage device and Internet in an efficient way and enables the usage of multimedia-enhanced remote education at home regardless of low speed transmission caused by narrow-bandwidth lines and modems. Possibilities to study at home, use multimedia courses without any limitations and have on-line connection to the teacher and colleagues open new directions in the educational process and will influence on improved and better education in the future.