Abstract | ||
---|---|---|
In this paper we propose a method for logo recognition using deep learning. Our recognition pipeline is composed of a logo region proposal followed by a Convolutional Neural Network (CNN) specifically trained for logo classification, even if they are not precisely localized. Experiments are carried out on the FlickrLogos-32 database, and we evaluate the effect on recognition performance of synthetic versus real data augmentation, and image pre-processing. Moreover, we systematically investigate the benefits of different training choices such as class-balancing, sample-weighting and explicit modeling the background class (i.e. no-logo regions). Experimental results confirm the feasibility of the proposed method, that outperforms the methods in the state of the art. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1016/j.neucom.2017.03.051 | Neurocomputing |
Keywords | DocType | Volume |
Logo recognition,Deep Learning,Convolutional Neural Network,Data augmentation,FlickrLogos-32 | Journal | abs/1701.02620 |
Issue | ISSN | Citations |
C | Neurocomputing 245, 23-30 (2017) | 14 |
PageRank | References | Authors |
0.74 | 22 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Simone Bianco | 1 | 226 | 24.48 |
Marco Buzzelli | 2 | 29 | 4.91 |
Davide Mazzini | 3 | 28 | 2.48 |
Raimondo Schettini | 4 | 1476 | 154.06 |