چکیده مقاله
Recent studies have made great progress in the transfer of style in the image and various methods have made progress in this field However, evaluating and improving the quality of stylization remain two important open challenges By examining these two aspects, in this paper, the style transfer quality is first decomposed into three measurable factors, namely Content Fidelity CF , Global Effects GE and Local Patterns LP Then, a new approach is further presented to exploit these factors to improve the quality of stylization The proposed approach is to use a deep neural network based on GAN or D GAN in general, which uses the manifold model
کلیدواژهها
نویسندگان
Mobin Khorushi
University of Mohaghegh Ardabili, Faculty of Technical Engineering, Ardabil, Iran
شیوه ارجاع
Khorushi, Mobin,1403,Image Quality Improving in the Field of Style Transfer based on Deep Learning,The 3th international conference on artificial intelligence and its future prospects in electrical, computer, mechanical and telecommunication engineering sciences.,Mashhad
ارائهشده در
مجموعه مقالات سومین کنفرانس بین المللی هوش مصنوعی و چشم انداز آینده آن در علوم مهندسی برق ، کامپیوتر ، مکانیک و مخابرات29 شهریور 1403 · مشهد