Noise Boosted Neural Receptive Fields

التفاصيل البيبلوغرافية
العنوان: Noise Boosted Neural Receptive Fields
المؤلفون: Federico Andrea Galatolo, Mario Giovanni Cosimo Antonio Cimino, Alessandro Marincioni, Gigliola Vaglini
المساهمون: Galatolo, FEDERICO ANDREA, Cimino, MARIO GIOVANNI COSIMO ANTONIO, Marincioni, Alessandro, Vaglini, Gigliola
بيانات النشر: IEEE Computer Society
USA
Washington
سنة النشر: 2021
المجموعة: ARPI - Archivio della Ricerca dell'Università di Pisa
مصطلحات موضوعية: neural network, supervised learning, boosting, receptive field, image classification
الوصف: Conventional neural networks (NNs) for image classification make use of a convolutional layer and a feedforward (FF) classification layer. This paper presents a novel classification layer architecture and a training paradigm, in which the FF layer is split into small and specialized FF nets called Noise Boosted Receptive Fields (NBRFs), one per class. Each i-th NBRF provides three membership degrees: to the i-th class, to the super class made by its complementary classes, and to an extra class representing out-of-classes images. The training process artificially generates extra-class samples, via image transformation and noise addition. Experimental results, carried out on MNIST, KMNIST and FMNIST datasets show that, with respect to an FF layer, the NBRF layer improves robustness and accuracy of classification. The repository with the source code and experimental data has been publicly released to facilitate reproducibility and widespread adoption.
نوع الوثيقة: conference object
وصف الملف: STAMPA
اللغة: English
العلاقة: info:eu-repo/semantics/altIdentifier/isbn/978-1-7281-9048-8; info:eu-repo/semantics/altIdentifier/wos/WOS:000824464300367; ispartofbook:2021 IEEE Symposium Series on Computational Intelligence (SSCI) Proceedings; 2021 IEEE Symposium Series on Computational Intelligence (SSCI); firstpage:1; lastpage:6; numberofpages:6; http://hdl.handle.net/11568/1122136Test; info:eu-repo/semantics/altIdentifier/scopus/2-s2.0-85125794662; https://ieeexplore.ieee.org/document/9660191Test
الإتاحة: http://hdl.handle.net/11568/1122136Test
https://ieeexplore.ieee.org/document/9660191Test
حقوق: info:eu-repo/semantics/openAccess
رقم الانضمام: edsbas.E74FE207
قاعدة البيانات: BASE