Hierarchical Adversarially Learned Inference

By Mohamed Ishmael Belghazi, Sai Rajeswar, Olivier Mastropietro, Negar Rostamzadeh, Jovana Mitrovi'c, Aaron C. Courville. 2018

29 citations recorded by the source corpus.

We propose a novel hierarchical generative model with a simple Markovian structure and a corresponding inference model. Both the generative and inference model are trained using the adversarial learning paradigm. We demonstrate that the hierarchical structure supports the learning of progressively more abstract representations as well as providing semantically meaningful reconstructions with different levels of fidelity. Furthermore, we show that minimizing the Jensen-Shanon divergence between the generative and inference network is enough to minimize the reconstruction error. The resulting semantically meaningful hierarchical latent structure discovery is exemplified on the CelebA dataset. There, we show that the features learned by our model in an unsupervised way outperform the best handcrafted features. Furthermore, the extracted features remain competitive when compared to several recent deep supervised approaches on an attribute prediction task on CelebA. Finally, we leverage the model's inference network to achieve state-of-the-art performance on a semi-supervised variant of the MNIST digit classification task.

Subfields: Generative Models

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Generative Models

Citations in the corpus

GAN Inversion: A Survey (2021, background): “Similar to GAN inversion, Adversarial Feature Learning (AFL, also known as BiGAN) [111], [175], [176] or Adversarially Learned Inference (ALI) [112], [177], [178] also aims to learn the inverse mapping of GAN models1.”
The challenge of realistic music generation: modelling raw audio at scale (2018, background): “Beyond raw audio generation, several generative models that capture large-scale structure have been proposed for images [2, 3, 12, 27], dialogue [40], 3D shapes [30] and symbolic music [37].”
Mutual Information Neural Estimation (2018, background): “Reconstructions are one desirable property of a model that does both inference and generation, but in practice ALI can lack ﬁdelity (i.e., reconstructs less faithfully than desired, see Li et al., 2017; Ulyanov et al., 2017; Belghazi et al., 2018).”
Augmented Normalizing Flows: Bridging the Gap Between Generative Flows and Latent Variable Models (2020, neutral): “Similar to other hierarchical models (Gulrajani et al., 2017b; Belghazi et al., 2018), ANF is also capable of retaining global, semantic information of the raw data stored in its higher level code; this is shown in Figure 7.”
TextKD-GAN: Text Generation Using Knowledge Distillation and Generative Adversarial Networks (2019, unknown): “AEs have been exploited along with GANs in diﬀerent architectures for computer vision application such as AAE [17], ALI [4], and HALI [1].”
Generative Adversarial Networks (GANs) (2020, methodology): “[87] attempts to extend ALI [83], named HALI, by achieving better perceptual matching in the reconstructions and by being capable to reduce the observables using a sequence of composed features maps.”