A carregar...
Publicação
Soundscape-sensing in social networks
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 757.2 KB | Adobe PDF |
Orientador(es)
Resumo(s)
In this paper we present the design and development stages of a social network mobile application based on soundscapes. The concept falls within the soundscape research and builds upon the assumption that soundscapes represent a valuable contribution for tracing the social context of a given place in a given time. This information is regarded as an important asset on social media interaction. The application includes four modules, which are minutely described in this paper: 1) A soundscape-sensing module that performs a classification task of computational auditory scene analysis, using Zero-Crossing and MFCC audio descriptors and a k-NN algorithm; 2) A client-server database module responsible for storing and distribute the resulting information among the peers of user’s social network. 3) A visualization module that shows clusters of peers based on their current acoustic similarities; 4) A web-based history module containing the evolution of users personal soundscapes. Preliminary results on the classification task indicate that simple classification (distinguishing between speech, music and environmental sounds) can be accomplished on a smartphone (iPhone). Energy resource tests show a considerable increase on battery usage, which may refrain users from use it continuously. Further tests will be accomplished meanwhile.
Descrição
Palavras-chave
Contexto Educativo
Citação
Cordeiro, J., Barbosa, Á., & Afonso, B. (2013). Soundscape-sensing in social networks. 198-199. Abstract from AIA-DAGA 2013, Merano, Italy.