Generative Adversarial Networks Based Reconstruction and Restoration of Cultural Heritage

الموضوعات

علوم تطبيقية وتكنولوجية

الملخص

Cultural heritage takes an important part in defining the identity and thehistory of a civilization or a nation. Valuing and preserving this heritage is thus a toppriority for governments and heritage institutions. Through this paper, we present animage completion (inpainting) approach adapted for the curation and the completionof damaged artwork. Our approach uses a set of machine learning techniques such asGenerative Adversarial Networks which are among the most powerful generativemodels that can be trained to generate realistic data samples. As we are focusingmostly on visual cultural heritage, the pipeline of our framework has manyoptimizations such as the use of clustering to optimize the training of the generativepart to ensure a better performance across a variety of cultural data categories. Theexperimental results of our framework were validated on cultural dataset of paintingscollected from Wiki-Art and the Rijksmuseum. We used the divide-and-conquerstrategy by clustering the training data into different small clusters containingsimilarly looking images to train smaller Specialized DCGANs. The training has beenmade on five painting categories containing 2000 paintings each, which took anaverage of 6.1 training hours. Training the Specialized DCGAN on 1200 paintingsfrom one of the clusters took 3.4 training hours. The inpainting results of theSpecialized DCGANs are clearly better in quality than the results of a DCGANtrained on mixture of paintings or on painting category.

روابط وملفات

التعريف والنوع

رقم الوثيقة

0c53415d-9d6f-40f9-a251-464cd2cac39b

رقم العقد

0

نوع الوسائط

Crawler

نوع المحتوى

الرسائل العلمية

صيغة المصدر

رسائل ماجيستير

نوع الملف

pdf text

أسماء الملفات

2018140_1.pdf

بيانات النشر

ألقاب المؤلفين

[{"name_ar":"عبدالعزيز بوراس","title_ar":"اشراف","title_en":"Supervision"},{"name_ar":"نسرين حمدالله جبور","title_ar":"اعداد","title_en":"Preparation"}]

اللغة

English

المصدر والدورية

اسم المصدر

Generative Adversarial Networks Based Reconstruction and Restoration of Cultural Heritage

المحتوى والصفحات

عدد الصفحات

0

إشراف وإعداد

الإشراف

عبدالعزيز بوراس

الإعداد

نسرين حمدالله جبور

الاقتباسات الببليوغرافية