A Neural Algorithm of Artistic Style by Leon A. Gatys, Alexander S. Ecker and Matthias Bethge
- How did they model “style” of an image so that they can learn the style and transfer it from one image to another?
- Representations of content and style in the CNN are separable.
- To generate image A from content of one image B and style of another image C, perform gradient descent on a white-noise image until content loss function (using layer 4 of VGG) is minimized between A and B and style loss function (using 5 layers of VGG) is minimized between A and C.
- content reconstruction vs. style reconstruction
- results generated on the basis of the VGG-Network, didn’t use FC layers
- replacing max-pooling with average pooling gave slightly more appealing results
- to perform content reconstruction, perform gradient descent on a white noise image to find another image that matches the feature responses of the original image
- Gram matrix and derivative notation was not fully defined, so I’m not sure what i, j, k represent
- Is gradient 0 if value is negative because they use ReLU?
- Is the assumption (for the algorithm to work) that multiple images can generate similar feature maps at higher levels of the CNN? And that multiple images can generate similar style representations?
- Did the algorithm differ from what you expected?
- Why are feature correlations representative of style?
- How might we improve on this algorithm?
- Performance: how long would it take to compute style transfer?
- How can we appply style transfer at B12?