TA2 (Together Anywhere, Together Anytime), pronounced "tattoo", aims at defining end-to-end systems for the development and delivery of new, creative forms of interactive, immersive, high quality media experiences for groups of users such as households and families. The overall vision of TA2 can be summarised as "making communications and engagement easier among groups of people separated in space and time".

One of the key components of TA2 is a set of generic and reliable audio, video, and multimodalities integration and recognition tools. This includes automatic extraction of cues from raw data streams. The running TA2 project stresses low-level "instantaneous" cues; it does not deal with semantic-aware integration of contextual information which could significantly improve the quality of cues.

The proposed TA2 project extension focuses on the medium-level (context-aware) cues taking into account not only low-level analysis outputs but also contextual information, e.g., about the activated scenario. The created semantic cues will be used by the TA2 system to orchestrate (i.e. frame, crop and represent) the audio-visual elements of the interaction between people.

The addition of BUT to the consortium will allow the semantic relevance of the metadata extracted from the analysis to be interpreted within the particular contexts described in the project. This will make the subsequent orchestration of the video more effective and more efficient and hence improve the end-user experience. The extension will enable building better applications that help families to interact easily and openly through games, through improved semi automatic production and publication of user generated content, and through enhanced ambient connectedness between families.

Description in Czech
TA2 (Together Anywhere, Together Anytime), pronounced "tattoo", aims at defining end-to-end systems for the development and delivery of new, creative forms of interactive, immersive, high quality media experiences for groups of users such as households and families. The overall vision of TA2 can be summarised as "making communications and engagement easier among groups of people separated in space and time".

One of the key components of TA2 is a set of generic and reliable audio, video, and multimodalities integration and recognition tools. This includes automatic extraction of cues from raw data streams. The running TA2 project stresses low-level "instantaneous" cues; it does not deal with semantic-aware integration of contextual information which could significantly improve the quality of cues.

The proposed TA2 project extension focuses on the medium-level (context-aware) cues taking into account not only low-level analysis outputs but also contextual information, e.g., about the activated scenario. The created semantic cues will be used by the TA2 system to orchestrate (i.e. frame, crop and represent) the audio-visual elements of the interaction between people.

The addition of BUT to the consortium will allow the semantic relevance of the metadata extracted from the analysis to be interpreted within the particular contexts described in the project. This will make the subsequent orchestration of the video more effective and more efficient and hence improve the end-user experience. The extension will enable building better applications that help families to interact easily and openly through games, through improved semi automatic production and publication of user generated content, and through enhanced ambient connectedness between families.

Keywords
social interaction, multimedia processing

7E11024

English

Smrž Pavel, doc. RNDr., Ph.D. - principal person responsible

Department of Computer Graphics and Multimedia
- responsible department (6.4.2010 - not assigned)

HRADIŠ, M.; BERAN, V.; ŘEZNÍČEK, I.; HEROUT, A.; BAŘINA, D.; VLČEK, A.; ZEMČÍK, P. Brno University of Technology at TRECVid 2010 SIN, CCD. In 2010 TREC Video Retrieval Evaluation Notebook Papers. Gaithersburg, MD: National Institute of Standards and Technology, 2010. p. 1-10.
Detail

ŽÁK, P.; BARTOŇ, R.; ZEMČÍK, P. Vision based user interface framework. Proceedings of the DT workshop. Žilina: 2010. p. 100.ISBN: 978-80-554-0304-5.
Detail

ŽÁK, P.; HRADIŠ, M.; SMRŽ, P.; ZEMČÍK, P.: Shared Image Preprocessing. URL: https://www.fit.vut.cz/research/product/209/. (Software)
Detail

HRADIŠ, M.; BAŘINA, D.; ŘEZNÍČEK, I.; ZEMČÍK, P.: Classifier creation framework for diverse classification tasks. URL: http://www.fit.vutbr.cz/research/prod/index.php.cs?id=155. (Software)
Detail

ŘEZNÍČEK, I.; HRADIŠ, M.; ZEMČÍK, P.: Online human action recognition framework. URL: http://www.fit.vutbr.cz/research/prod/index.php.cs?id=204. (Software)
Detail

ŘEZNÍČEK, I.; ZEMČÍK, P. On-line human action detection using space-time interest points. Zborník príspevkov prezentovaných na konferencii ITAT, september 2011. Praha: Faculty of Mathematics and Physics, Charles University, 2011. p. 39.ISBN: 978-80-89557-01-1.
Detail

ŘEZNÍČEK, I.; BAŘINA, D. Classifier creation framework for diverse classification tasks. Proceedings of the DT workshop. Žilina: Brno University of Technology, 2010. p. 50.ISBN: 978-80-554-0304-5.
Detail

HRADIŠ, M.; ŘEZNÍČEK, I.; BEHÚŇ, K. Brno University of Technology at MediaEval 2011 Genre Tagging Task. Working Notes Proceedings of the MediaEval 2011 Workshop. CEUR workshop proceedings. Pisa, Italy: CEUR-WS.org, 2011. iss. 9, p. 1.ISSN: 1613-0073.
Detail

BEDNAŘÍK, R.; VRZÁKOVÁ, H.; HRADIŠ, M. What you want to do next: A novel approach for intent prediction in gaze-based interaction. ETRA '12 Proceedings of the Symposium on Eye Tracking Research and Applications. Santa Barbara: Association for Computing Machinery, 2012. p. 83.ISBN: 978-1-4503-1221-9.
Detail

HRADIŠ, M.; ŘEZNÍČEK, I.; BEHÚŇ, K. Semantic Class Detectors in Video Genre Recognition. Proceedings of VISAPP 2012. Rome: SciTePress - Science and Technology Publications, 2012. p. 640.ISBN: 978-989-8565-03-7.
Detail

HRADIŠ, M.; EIVAZI, S.; BEDNAŘÍK, R. Voice activity detection in video mediated communication from gaze. ETRA '12 Proceedings of the Symposium on Eye Tracking Research and Applications. Santa Barbara: Association for Computing Machinery, 2012. p. 329.ISBN: 978-1-4503-1221-9.
Detail

MOTLÍČEK, P.; VALENTE, F.; SZŐKE, I. Improving Acoustic Based Keyword Spotting Using LVCSR Lattices. Proc. International Conference on Acoustics, Speech, and Signal Processing 2012. Kyoto: IEEE Signal Processing Society, 2012. p. 4413.ISBN: 978-1-4673-0044-5.
Detail

POLÁČEK, O.; KLÍMA, M.; SPORKA, A.; ŽÁK, P.; HRADIŠ, M.; ZEMČÍK, P.; PROCHÁZKA, V. A Comparative Study on Distant Free-Hand Pointing. EuroiTV '12 Proceedings of the 10th European conference on Interactive tv and video. Berlin, Germany: Association for Computing Machinery, 2012. p. 139.ISBN: 978-1-4503-1107-6.
Detail

KRÁL, J.; HRADIŠ, M. Restricted Boltzman Machines for Image Tag Suggestion. Proceedings of the 19th Conference STUDENT EEICT 2012. Brno: Brno University of Technology, 2012. p. 1.
Detail

Responsibility: Smrž Pavel, doc. RNDr., Ph.D.