| | *ISCA SIG on Speech and Language Indexing for Multimedia*
We would like to invite everyone involved in the field of automatic speech and language processing on multimedia data to participate in the creation of a Speech and Language Indexing for Multimedia (SLIM) ISCA Special Interest Group (SIG). To push forward this proposal and better define the expectations and missions for such a SIG, we invite you to a meeting at Interspeech, on Wed. 31 August 12:00-12:45 (during lunch break). The meeting place will be announced at Interspeech.
The SLIM SIG intends to put together all protagonists working on speech and language processing to analyze, index and access multimedia data. Spoken language clearly plays a major role in semantic access to multimedia data such as lectures, meetings, interviews, debates, conversational broadcast, podcasts, social videos on the Web, etc. However, such data raise specific challenges to speech technologies: high transcription error rates can be observed due to the sometime poor quality of the data; semantics can be shared between several modalities; etc. Bringing together researchers from across our community will allow us to direct focused effort at addressing these challenges. The initiative is of interest to people working in areas related to (but not limited to): Spoken content retrieval and spoken term detection for multimedia collections, speech summarization, speech-aware multimedia search engines and automatic speech structuring for indexing purposes, speech-based content recommendation and archive mining. The SLIM SIG offers a unique opportunity to share experience in multimedia spoken data processing, to foster new research directions and to gain visibility both in the speech and multimedia community. To do so, several catalysts can be foreseen: animate a discussion forum and a web site, provide a directory of resources and data, promote evaluation campaigns on multimedia data (initiatives such as MediaEval, ESTER/ETAPE, as well as many others), organize special sessions in the main speech and multimedia conferences, organize dedicated workshops, build strong relations with SIGs in other communities (IEEE, ACM), etc.
Feel free to contact us by mail should not be able to attend the meeting in Florence.
Guillaume Gravier (guig@irisa.fr), CNRS-IRISA, France
Martha Larson (m.a.larson@tudelft.nl), TU Delft, The Netherlands
Gareth Jones (gareth.jones@computing.dcu.ie), DCU, Ireland |
