Fusarium Damaged Kernels Detection Using Transfer Learning on Deep Neural Network Architecture. Nicolau, M., Pimentel, M. B. M., Tibola, C. S., Fernandes, J. M. C., & Pavan, W. abstract bibtex The present work shows the application of transfer learning for a pre-trained deep neural network (DNN), using a small image dataset (≈ 12,000) on a single workstation with enabled NVIDIA GPU card that takes up to 1 hour to complete the training task and archive an overall average accuracy of 94.7%. The DNN presents a 20% score of misclassification for an external test dataset. The accuracy of the proposed methodology is equivalent to ones using HSI methodology (81%-91%) used for the same task, but with the advantage of being independent on special equipment to classify wheat kernel for FHB symptoms.
@article{nicolau_fusarium_nodate,
title = {Fusarium {Damaged} {Kernels} {Detection} {Using} {Transfer} {Learning} on {Deep} {Neural} {Network} {Architecture}},
abstract = {The present work shows the application of transfer learning for a pre-trained deep neural network (DNN), using a small image dataset (≈ 12,000) on a single workstation with enabled NVIDIA GPU card that takes up to 1 hour to complete the training task and archive an overall average accuracy of 94.7\%. The DNN presents a 20\% score of misclassification for an external test dataset. The accuracy of the proposed methodology is equivalent to ones using HSI methodology (81\%-91\%) used for the same task, but with the advantage of being independent on special equipment to classify wheat kernel for FHB symptoms.},
language = {en},
author = {Nicolau, Marcio and Pimentel, Marcia Barrocas Moreira and Tibola, Casiane Salete and Fernandes, Jose Mauricio Cunha and Pavan, Willingthon},
pages = {9}
}
Downloads: 0
{"_id":"xLnyWQfSjxFRybwQY","bibbaseid":"nicolau-pimentel-tibola-fernandes-pavan-fusariumdamagedkernelsdetectionusingtransferlearningondeepneuralnetworkarchitecture","downloads":0,"creationDate":"2019-04-04T09:48:13.774Z","title":"Fusarium Damaged Kernels Detection Using Transfer Learning on Deep Neural Network Architecture","author_short":["Nicolau, M.","Pimentel, M. B. M.","Tibola, C. S.","Fernandes, J. M. C.","Pavan, W."],"year":null,"bibtype":"article","biburl":"https://api.zotero.org/users/2700485/collections/GNU8E8T8/items?key=mDatGbIIG9UB5rvnQ0sLpyWn&format=bibtex&limit=100","bibdata":{"bibtype":"article","type":"article","title":"Fusarium Damaged Kernels Detection Using Transfer Learning on Deep Neural Network Architecture","abstract":"The present work shows the application of transfer learning for a pre-trained deep neural network (DNN), using a small image dataset (≈ 12,000) on a single workstation with enabled NVIDIA GPU card that takes up to 1 hour to complete the training task and archive an overall average accuracy of 94.7%. The DNN presents a 20% score of misclassification for an external test dataset. The accuracy of the proposed methodology is equivalent to ones using HSI methodology (81%-91%) used for the same task, but with the advantage of being independent on special equipment to classify wheat kernel for FHB symptoms.","language":"en","author":[{"propositions":[],"lastnames":["Nicolau"],"firstnames":["Marcio"],"suffixes":[]},{"propositions":[],"lastnames":["Pimentel"],"firstnames":["Marcia","Barrocas","Moreira"],"suffixes":[]},{"propositions":[],"lastnames":["Tibola"],"firstnames":["Casiane","Salete"],"suffixes":[]},{"propositions":[],"lastnames":["Fernandes"],"firstnames":["Jose","Mauricio","Cunha"],"suffixes":[]},{"propositions":[],"lastnames":["Pavan"],"firstnames":["Willingthon"],"suffixes":[]}],"pages":"9","bibtex":"@article{nicolau_fusarium_nodate,\n\ttitle = {Fusarium {Damaged} {Kernels} {Detection} {Using} {Transfer} {Learning} on {Deep} {Neural} {Network} {Architecture}},\n\tabstract = {The present work shows the application of transfer learning for a pre-trained deep neural network (DNN), using a small image dataset (≈ 12,000) on a single workstation with enabled NVIDIA GPU card that takes up to 1 hour to complete the training task and archive an overall average accuracy of 94.7\\%. The DNN presents a 20\\% score of misclassification for an external test dataset. The accuracy of the proposed methodology is equivalent to ones using HSI methodology (81\\%-91\\%) used for the same task, but with the advantage of being independent on special equipment to classify wheat kernel for FHB symptoms.},\n\tlanguage = {en},\n\tauthor = {Nicolau, Marcio and Pimentel, Marcia Barrocas Moreira and Tibola, Casiane Salete and Fernandes, Jose Mauricio Cunha and Pavan, Willingthon},\n\tpages = {9}\n}\n\n","author_short":["Nicolau, M.","Pimentel, M. B. M.","Tibola, C. S.","Fernandes, J. M. C.","Pavan, W."],"key":"nicolau_fusarium_nodate","id":"nicolau_fusarium_nodate","bibbaseid":"nicolau-pimentel-tibola-fernandes-pavan-fusariumdamagedkernelsdetectionusingtransferlearningondeepneuralnetworkarchitecture","role":"author","urls":{},"downloads":0},"search_terms":["fusarium","damaged","kernels","detection","using","transfer","learning","deep","neural","network","architecture","nicolau","pimentel","tibola","fernandes","pavan"],"keywords":[],"authorIDs":[],"dataSources":["WjaEGHGBpByAsAbzq"]}