Sound Research WIKINDX

WIKINDX Resources

Cano, P., Koppenberger, M., le Groux, S., Ricard, J., Wack, N., & Herrera, P. (2005). Nearest neighbor automatic sound annotation with a WordNet taxonomy. Journal of Intelligent Systems, 24(2/3), 99–111. 
Added by: Mark Grimshaw-Aagaard (1/26/06, 11:52 AM)   Last edited by: Mark Grimshaw-Aagaard
Resource type: Journal Article
BibTeX citation key: Cano2005
Email resource to friend
View all bibliographic details
Categories: General, Typologies/Taxonomies
Keywords: Semantic categorization, Synchresis/Synchrony
Creators: Cano, le Groux, Herrera, Koppenberger, Ricard, Wack
Collection: Journal of Intelligent Systems
Resources citing this (Bibliography: WIKINDX Master Bibliography)
Views: 14/1148

Experimental results for a general sound annotator allowing for the selection of sound [FX] by sound [property] categorization rather than text [descriptive] selection.

See also (Xu et al. 2004; Khan, McLeod, & Hovy 2004)

Khan, L., McLeod, D., & Hovy, E. (2004). Retrieval effectiveness of an ontology-based model for information selection. Very Large Data Bases, 13, 71–85.
Xu, M., Duan, L.-Y., Cai, J., Chia, L.-T., Xu, C., & Tian, Q. (2004). HMM-based audio keyword generation. Lecture Notes in Computer Science, 3333, 556–574.
Added by: Mark Grimshaw-Aagaard  Last edited by: Mark Grimshaw-Aagaard
p.100   Summarizing other research, states that "one of the main problems faced by natural sounds and sound effects classifiers is the lack of a clear taxonomy." For musical instruments "there is a parallelism between semantic and perceptual taxonomies" that does not exist for "every-day sound classification".

e.g. musical instruments are sustained, not sustained, string, brass etc. However, "one can find hissing sounds in categories of "cat", "tea boilers', "snakes." Foley artists exploit this ambiguity" by the use of sounds unconnected with a visual object.   Added by: Mark Grimshaw-Aagaard