accelbrainbase.observabledata._mxnet.adversarialmodel.generativemodel package

Submodules

accelbrainbase.observabledata._mxnet.adversarialmodel.generativemodel.convolutional_auto_encoder module

class accelbrainbase.observabledata._mxnet.adversarialmodel.generativemodel.convolutional_auto_encoder.ConvolutionalAutoEncoder(noise_sampler, model, initializer=None, condition_sampler=None, conditonal_dim=1, learning_rate=1e-05, optimizer_name='SGD', hybridize_flag=True, scale=1.0, ctx=gpu(0), **kwargs)

Bases: accelbrainbase.observabledata._mxnet.adversarialmodel.generative_model.GenerativeModel

Generative model, which draws samples from the fake distribution, in the Generative Adversarial Networks(GANs).

The Generative Adversarial Networks(GANs) (Goodfellow et al., 2014) framework establishes a min-max adversarial game between two neural networks – a generative model, G, and a discriminative model, D. The discriminator model, D(x), is a neural network that computes the probability that a observed data point x in data space is a sample from the data distribution (positive samples) that we are trying to model, rather than a sample from our generative model (negative samples).

Concurrently, the generator uses a function G(z) that maps samples z from the prior p(z) to the data space. G(z) is trained to maximally confuse the discriminator into believing that samples it generates come from the data distribution. The generator is trained by leveraging the gradient of D(x) w.r.t. x, and using that to modify its parameters.

References

  • Gauthier, J. (2014). Conditional generative adversarial nets for convolutional face generation. Class Project for Stanford CS231N: Convolutional Neural Networks for Visual Recognition, Winter semester, 2014(5), 2.
  • Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., … & Bengio, Y. (2014). Generative adversarial nets. In Advances in neural information processing systems (pp. 2672-2680).
  • Makhzani, A., Shlens, J., Jaitly, N., Goodfellow, I., & Frey, B. (2015). Adversarial autoencoders. arXiv preprint arXiv:1511.05644.
  • Mirza, M., & Osindero, S. (2014). Conditional generative adversarial nets. arXiv preprint arXiv:1411.1784.
  • Salimans, T., Goodfellow, I., Zaremba, W., Cheung, V., Radford, A., & Chen, X. (2016). Improved techniques for training gans. In Advances in neural information processing systems (pp. 2234-2242).
  • Zhao, J., Mathieu, M., & LeCun, Y. (2016). Energy-based generative adversarial network. arXiv preprint arXiv:1609.03126.
  • Warde-Farley, D., & Bengio, Y. (2016). Improving generative adversarial networks with denoising feature matching.
draw()

Draw samples from the fake distribution.

Returns:Tuple of `mxnet.ndarray`s.
get_model()

getter for mxnet.gluon.hybrid.hybridblock.HybridBlock.

model

getter for mxnet.gluon.hybrid.hybridblock.HybridBlock.

set_model(value)

setter for mxnet.gluon.hybrid.hybridblock.HybridBlock.

Module contents