Title | Neural Style Transfer for Picture with Gradient Gram Matrix Description |
Publication Type | Conference Paper |
Year of Publication | 2020 |
Authors | Jin, H., Wang, T., Zhang, M., Li, M., Wang, Y., Snoussi, H. |
Conference Name | 2020 39th Chinese Control Conference (CCC) |
Keywords | art style transfer, Automation, electrical engineering, Gradient Gram description, Gradient Gram matrix, gradient gram matrix description, Image reconstruction, image texture, Internet of Things, matrix algebra, muddiness, neural style transfer, output stylized picture, picture stylize, Predictive Metrics, pubcrawl, Resiliency, Runtime, Scalability, style loss, stylized pictures, texture details, texture expression, texture style, US Department of Defense |
Abstract | Despite the high performance of neural style transfer on stylized pictures, we found that Gatys et al [1] algorithm cannot perfectly reconstruct texture style. Output stylized picture could emerge unsatisfied unexpected textures such like muddiness in local area and insufficient grain expression. Our method bases on original algorithm, adding the Gradient Gram description on style loss, aiming to strengthen texture expression and eliminate muddiness. To some extent our method lengthens the runtime, however, its output stylized pictures get higher performance on texture details, especially in the elimination of muddiness. |
DOI | 10.23919/CCC50068.2020.9188652 |
Citation Key | jin_neural_2020 |