Visual recognition using artificial intelligence travel image generation using neural style transfer

• Main Author: Liu, Tianshu
• Other Authors: Yap Kim Hui
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/139883
• Institution: Nanyang Technological University

With the continuous improvement of the life quality, more and more people start to travel abroad during their holidays. There is no denying that people spend a lot of time buying souvenirs for ourselves and most important, leaving us a special memory. However, people are no longer satisfied with the gifts that follow the same pattern. This Travel Image Generation aims to abstract the art style from the famous local painting or photo that people take during their trips and generate the unique travel photos that only belonging to themselves. A beard well lathered is half shaved. The first part of this project focus on the new acknowledges learning. Related literature and lecture notes have been reviewed. Software/library installation and GPU have been well prepared. The second part of this project is to explore the network of the system. Here we used StyleBank [1]. StyleBank consists of multiple convolution filter banks while each filter bank represents one style accordingly. Stylization results will be generated by mapping the StyleBank to the content image. Lastly, the performance of the networks will be tested. We also discussed the difference between stylization results between landscape and portrait. The possible improvement for the whole system will also be mentioned in the report.