@article{bianco_deep_2017,
	title = {Deep learning for logo recognition},
	volume = {245},
	issn = {0925-2312},
	url = {http://www.sciencedirect.com/science/article/pii/S0925231217305660},
	doi = {10.1016/j.neucom.2017.03.051},
	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.},
	urldate = {2018-03-25TZ},
	journal = {Neurocomputing},
	author = {Bianco, Simone and Buzzelli, Marco and Mazzini, Davide and Schettini, Raimondo},
	month = jul,
	year = {2017},
	keywords = {Convolutional Neural Network, Data augmentation, Deep Learning, FlickrLogos-32, Logo recognition},
	pages = {23--30}
}

{"_id":"DkT6HGLnbYRNTYtGi","bibbaseid":"bianco-buzzelli-mazzini-schettini-deeplearningforlogorecognition-2017","downloads":0,"creationDate":"2018-04-15T14:41:43.968Z","title":"Deep learning for logo recognition","author_short":["Bianco, S.","Buzzelli, M.","Mazzini, D.","Schettini, R."],"year":2017,"bibtype":"article","biburl":"https://bibbase.org/zotero/pvhuy","bibdata":{"bibtype":"article","type":"article","title":"Deep learning for logo recognition","volume":"245","issn":"0925-2312","url":"http://www.sciencedirect.com/science/article/pii/S0925231217305660","doi":"10.1016/j.neucom.2017.03.051","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.","urldate":"2018-03-25TZ","journal":"Neurocomputing","author":[{"propositions":[],"lastnames":["Bianco"],"firstnames":["Simone"],"suffixes":[]},{"propositions":[],"lastnames":["Buzzelli"],"firstnames":["Marco"],"suffixes":[]},{"propositions":[],"lastnames":["Mazzini"],"firstnames":["Davide"],"suffixes":[]},{"propositions":[],"lastnames":["Schettini"],"firstnames":["Raimondo"],"suffixes":[]}],"month":"July","year":"2017","keywords":"Convolutional Neural Network, Data augmentation, Deep Learning, FlickrLogos-32, Logo recognition","pages":"23--30","bibtex":"@article{bianco_deep_2017,\n\ttitle = {Deep learning for logo recognition},\n\tvolume = {245},\n\tissn = {0925-2312},\n\turl = {http://www.sciencedirect.com/science/article/pii/S0925231217305660},\n\tdoi = {10.1016/j.neucom.2017.03.051},\n\tabstract = {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.},\n\turldate = {2018-03-25TZ},\n\tjournal = {Neurocomputing},\n\tauthor = {Bianco, Simone and Buzzelli, Marco and Mazzini, Davide and Schettini, Raimondo},\n\tmonth = jul,\n\tyear = {2017},\n\tkeywords = {Convolutional Neural Network, Data augmentation, Deep Learning, FlickrLogos-32, Logo recognition},\n\tpages = {23--30}\n}\n\n","author_short":["Bianco, S.","Buzzelli, M.","Mazzini, D.","Schettini, R."],"key":"bianco_deep_2017","id":"bianco_deep_2017","bibbaseid":"bianco-buzzelli-mazzini-schettini-deeplearningforlogorecognition-2017","role":"author","urls":{"Paper":"http://www.sciencedirect.com/science/article/pii/S0925231217305660"},"keyword":["Convolutional Neural Network","Data augmentation","Deep Learning","FlickrLogos-32","Logo recognition"],"downloads":0},"search_terms":["deep","learning","logo","recognition","bianco","buzzelli","mazzini","schettini"],"keywords":["convolutional neural network","data augmentation","deep learning","flickrlogos-32","logo recognition"],"authorIDs":[],"dataSources":["E2tDnyqghkeNYrDGK"]}