Abstract
Animal counting is often employed to estimate the population size and species diversity in a given area. This work makes a start at the unsupervised segregation of audio clips with a Self Organizing Map (SOM), using Speeded Up Robust Features (SURF) applied to audio spectrograms as inputs. Further work is required for a functional algorithm.
| Original language | English |
|---|---|
| Title of host publication | 2016 13th Symposium on Neural Networks and Applications (NEUREL) : November 22-24, 2016, SAVA Center, Milentij a Popovica 9, Belgrade, Serbia |
| Editors | S. Relijn, S. Stanković |
| Place of Publication | Piscataway |
| Publisher | Institute of Electrical and Electronics Engineers |
| ISBN (Electronic) | 9781509015306 |
| ISBN (Print) | 978-1-5090-1529-0 |
| DOIs | |
| Publication status | Published - 27 Dec 2016 |
| Event | 13th Symposium on Neural Networks and Applications (NEUREL 2016) - SAVA Center, Belgrade, Serbia Duration: 22 Nov 2016 → 24 Nov 2016 Conference number: 13 |
Conference
| Conference | 13th Symposium on Neural Networks and Applications (NEUREL 2016) |
|---|---|
| Abbreviated title | NEUREL 2016 |
| Country/Territory | Serbia |
| City | Belgrade |
| Period | 22/11/16 → 24/11/16 |
Keywords
- Self Organizing Map (SOM)
- Speeded Up Robust Features (SURF)
- Unsupervised audio segregation