Perceptual Losses for Real-Time Style Transfer and Super-Resolution
Citations
11,127 citations
5,782 citations
Cites background from "Perceptual Losses for Real-Time Sty..."
...[35] is much faster, but limits transfer to a pre-trained set of styles....
[...]
4,465 citations
Cites background from "Perceptual Losses for Real-Time Sty..."
...[21] who have shown impressive results for neural style transfer and superresolution....
[...]
...Neural Style Transfer [11, 21, 48, 10] is another way to perform image-to-image translation, which synthesizes a novel image by combining the content of one image with the style of another image (typically a painting) by matching the Gram matrix statistics of pre-trained deep features....
[...]
...[21], we use instance normalization [49]....
[...]
...Neural Style Transfer [12, 22, 51, 11] is another way to perform image-to-image translation, which synthesizes a novel image by combining the content of one image with the style of another image (typically a painting) based on matching the Gram matrix statistics of pre-trained deep features....
[...]
4,404 citations
Cites background or methods from "Perceptual Losses for Real-Time Sty..."
...The MOS scores obtained with SRGAN are closer to those of the original high-resolution images than to those obtained with any state-of-the-art method....
[...]
...This is illustrated in Figure 2, where highest PSNR does not necessarily reflect the perceptually better SR result....
[...]
...Prediction-based methods were among the first methods to tackle SISR....
[...]
...This is an improvement over Dong et al. [9] where bicubic interpolation is employed to upscale the LR observation before feeding the image to the CNN....
[...]
3,838 citations
Cites methods from "Perceptual Losses for Real-Time Sty..."
...These methods measure distance in VGG feature space as a “perceptual loss” for image regression problems [23, 14]....
[...]
...For example, features from the VGG architecture [51] have been used on tasks such as neural style transfer [17], image superresolution [23], and conditional image synthesis [14, 8]....
[...]
References
123,388 citations
111,197 citations
"Perceptual Losses for Real-Time Sty..." refers methods in this paper
...We train with a batch size of 4 for 200k iterations using Adam [51] with a learning rate of 1×10−3 without weight decay or dropout....
[...]
...a batch size of 4 for 200k iterations using Adam [51] with a learning rate of 1×10−3 without weight decay or dropout....
[...]
...We use Adam [51] with a learning rate of 1× 10−3....
[...]
55,235 citations
"Perceptual Losses for Real-Time Sty..." refers methods in this paper
...In all our experiments φ is the 16-layer VGG network [46] pretrained on the ImageNet dataset [47]....
[...]
49,914 citations
40,609 citations