11:05 AM - 11:30 AM
[AIS3/VHF5-2(Invited)] Improving Material Translation Based on Style Image Retrieval for Neural Style Transfer
style image retrieval, material translation, neural style transfer
In this talk, we introduce a CNN-feature-based image retrieval method to find the ideal style image that better translates the material of an object. We segment objects from the content image by using a weakly supervised segmentation method, and transfer the material of the retrieved style image to the segmented areas. With this method, we achieve realistic images that can even fool human perception.