Re:Playing Around with Noise as Targets
by DoctorOetker on Friday August 10, 2018 @07:30AM
(#17732351)
I didn't read the referenced 2017 paper yet, but mapping the training data to noise (gaussian and/or other) is exactly what the RevNet paper does, with the advantage of deterministic reversibility such that the trained RevNet is also generative (without having to do gradient descent for each generated image)
Re:Playing Around with Noise as Targets
by dtjohnnyb on Friday August 10, 2018 @08:41AM
(#17732717)
The intro to the paper has a nice comparison to other similar methods (generative and non-generative) and the blog post linked in this article by inFERNCe https://www.inference.vc/unsupervised-learning-by-predicting... has a nice comparison at the end to different unsupervised methods and where this method adds novelty (or doesn't!)
Re:Playing Around with Noise as Targets
by cgearhart on Friday August 10, 2018 @09:25AM
(#17732973)
His stated purpose was image compression (although I didn’t see evidence that it worked). If the distribution encoded by the is smaller than your image, then you can send the small set of model parameters (instead of the image) and then use the model to reconstruct the target image.
I don't see the comparisons at the end of the inFERENCe link?
A visual proof that neural nets can compute any function: http://neuralnetworksanddeeplearning.com/chap4.html
I fear those words