This paper aims to combine intuition and practical experience in the context of ImageCLEF 2013 Plant Identification task.

We propose a flexible, modular system which allows us to analyse and combine the results after apply- ing methods such as image retrieval using LIRe, metadata clustering and naive Bayes classification. Although the training collection is quite extensive, cover- ing a large number of species, in order to obtain accurate results with our photo annotation algorithm we enriched our system with new images from a reliable source.

As a result, we performed four runs with different configurations, and the best run was ranked 5 th out of a total of 12 group participants.

Questo è uno degli articoli scientifici pubblicati da uno o più collaboratori e data scientist di synbrAIn.

Se sei interessato a saperne di più, leggi l’intero articolo qui.