The original idea and algorithm of using a deep neural network to merge the content of an image with the style of another was published in a paper titled A Neural Algorithm of Artistic Style (https://arxiv.org/abs/1508.06576) in the summer of 2015. It was based on a pre-trained deep CNN model called VGG-19 (https://arxiv.org/pdf/1409.1556.pdf), the winner of the 2014 ImageNet image recognition challenge, which has 16 convolutional layers, or feature maps, representing different levels of the image content. In this original method, the final transferred image is first initialized as a white noise image merged with the content image. The content loss function is defined as the squared error loss of a specific set of feature representations on the convolutional layer, conv4_2, of the content image and the result image after both being...
United States
Great Britain
India
Germany
France
Canada
Russia
Spain
Brazil
Australia
Singapore
Hungary
Ukraine
Luxembourg
Estonia
Lithuania
South Korea
Turkey
Switzerland
Colombia
Taiwan
Chile
Norway
Ecuador
Indonesia
New Zealand
Cyprus
Denmark
Finland
Poland
Malta
Czechia
Austria
Sweden
Italy
Egypt
Belgium
Portugal
Slovenia
Ireland
Romania
Greece
Argentina
Netherlands
Bulgaria
Latvia
South Africa
Malaysia
Japan
Slovakia
Philippines
Mexico
Thailand