2016

Task Specific Adversarial Cost Function

Creswell, Antonia, Bharath, Anil A.

Understand

The cost function used to train a generative model should fit the purpose of the model.

If the model is intended for tasks such as generating perceptually correct samples, it is beneficial to maximise the likelihood of a sample drawn from the model, Q, coming from the same distribution as the training data, P.
This is equivalent to minimising the Kullback-Leibler (KL) distance, KL[Q||P].
However, if the model is intended for tasks such as retrieval or classification it is beneficial to maximise the likelihood that a sample drawn from the training data is captured by the model, equivalent to minimising KL[P||Q].

“Animating rotation with quaternion curves”
Ken Shoemake · 1985
Earlier work this paper cites.
“Divergence measures based on the Shannon entropy”
Jianhua Lin · 1991
Earlier work this paper cites.
“Extracting and composing robust features with denoising autoencoders”
Pascal Vincent, Hugo Larochelle, Yoshua Bengio and Pierre-Antoine Manzagol · 2008
Earlier work this paper cites.
“Why does unsupervised pre-training help deep learning?”
Dumitru Erhan et al · 2010
Earlier work this paper cites.
“Imagenet classification with deep convolutional neural networks”
Alex Krizhevsky, Ilya Sutskever and Geoffrey˜E Hinton · 2012
Earlier work this paper cites.
“Generalized denoising auto-encoders as generative models”
Yoshua Bengio, Li Yao, Guillaume Alain and Pascal Vincent · 2013
Earlier work this paper cites.
“Domain-adversarial neural networks”
Original
Hana Ajakan et al · 2014
Earlier work this paper cites.
“Conditional generative adversarial nets for convolutional face generation”
Jon Gauthier · 2014
Earlier work this paper cites.
“Generative adversarial nets”
Ian Goodfellow et al · 2014
Earlier work this paper cites.
“Auto-Encoding Variational Bayes”
Diederik˜P Kingma and Max Welling · 2014
Earlier work this paper cites.
“Conditional generative adversarial nets”
Original
Mehdi Mirza and Simon Osindero · 2014
Earlier work this paper cites.

“Learning to generate chairs with convolutional neural networks”
Alexey Dosovitskiy, Jost Tobias˜Springenberg and Thomas Brox · 2015
Cited alongside, same era.
“DRAW: A Recurrent Neural Network For Image Generation”
Karol Gregor et al · 2015
Cited alongside, same era.
“How (not) to Train your Generative Model: Scheduled Sampling, Likelihood, Adversary?”
Original
Ferenc Husz“’ar · 2015
Cited alongside, same era.
“Human-level concept learning through probabilistic program induction”
Brenden˜M Lake, Ruslan Salakhutdinov and Joshua˜B Tenenbaum · 2015
Cited alongside, same era.
“Fully convolutional networks for semantic segmentation”
Jonathan Long, Evan Shelhamer and Trevor Darrell · 2015
Cited alongside, same era.
“Going deeper with convolutions”
Christian Szegedy et al · 2015
Cited alongside, same era.
“Sketch-a-net that beats humans”
Qian Yu et al · 2015
Cited alongside, same era.
“InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets”
Xi Chen et al · 2016
Cited alongside, same era.
“Adversarial Training For Sketch Retrieval”
Antonia Creswell and Anil˜Anthony Bharath · 2016
Cited alongside, same era.
“Adversarial Feature Learning”
Original
Jeff Donahue, Philipp Kr“”ahenb“”uhl and Trevor Darrell · 2016
Cited alongside, same era.
“Adversarially Learned Inference”
Original
Vincent Dumoulin et al · 2016
Cited alongside, same era.

“Domain-adversarial training of neural networks”
Yaroslav Ganin et al · 2016
Closest in time.
“Image generation using generative adversarial networks and attention mechanism”
Y. Kataoka, T. Matsubara and K. Uehara · 2016
Closest in time.
“Autoencoding beyond pixels using a learned similarity metric”
Anders Boesen˜Lindbo Larsen, Sren˜Kaae Snderby and Ole Winther · 2016
Closest in time.
“Adversarial Autoencoders”
Original
Alireza Makhzani, Jonathon Shlens, Navdeep Jaitly and Ian Goodfellow · 2016
Closest in time.
“Pixel recurrent neural networks”
Aaron van˜den Oord, Nal Kalchbrenner and Koray Kavukcuoglu · 2016
Closest in time.
“Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks”
Alec Radford, Luke Metz and Soumith Chintala · 2016
Closest in time.
“Improved Techniques for Training GANs”
Tim Salimans et al · 2016
Closest in time.
“One-shot Learning with Memory-Augmented Neural Networks”
Original
Adam Santoro et al · 2016
Closest in time.
“A note on the evaluation of generative models” International Conference on Learning Representations, 2016
Original
L. Theis, A. Oord and M. Bethge · 2016
Closest in time.
“Matching Networks for One Shot Learning”
Oriol Vinyals et al · 2016
Closest in time.
“Energy-based Generative Adversarial Network”
Original
Junbo Zhao, Michael Mathieu and Yann LeCun · 2016
Closest in time.

alphaXiv searches the wider corpus for related work and actual follow-ups.

alphaXiv is searching for related work…

Reading the bibliography…