Viet-Anh on Software Logo

What is: VGG Loss?

SourcePhoto-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
Year2000
Data SourceCC BY-SA - https://paperswithcode.com

VGG Loss is a type of content loss introduced in the Perceptual Losses for Real-Time Style Transfer and Super-Resolution super-resolution and style transfer framework. It is an alternative to pixel-wise losses; VGG Loss attempts to be closer to perceptual similarity. The VGG loss is based on the ReLU activation layers of the pre-trained 19 layer VGG network. With ϕ_i,j\phi\_{i,j} we indicate the feature map obtained by the jj-th convolution (after activation) before the ii-th maxpooling layer within the VGG19 network, which we consider given. We then define the VGG loss as the euclidean distance between the feature representations of a reconstructed image G_θ_G(ILR)G\_{\theta\_{G}}\left(I^{LR}\right) and the reference image IHRI^{HR}:

l_VGG/i.j=1W_i,jH_i,j_x=1W_i,j_y=1H_i,j(ϕ_i,j(IHR)_x,yϕ_i,j(G_θ_G(ILR))_x,y)2 l\_{VGG/i.j} = \frac{1}{W\_{i,j}H\_{i,j}}\sum\_{x=1}^{W\_{i,j}}\sum\_{y=1}^{H\_{i,j}}\left(\phi\_{i,j}\left(I^{HR}\right)\_{x, y} - \phi\_{i,j}\left(G\_{\theta\_{G}}\left(I^{LR}\right)\right)\_{x, y}\right)^{2}

Here W_i,jW\_{i,j} and H_i,jH\_{i,j} describe the dimensions of the respective feature maps within the VGG network.