The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Le modèle graphique profond (DGM) basé sur les réseaux contradictoires génératifs (GAN) s'est révélé prometteur en matière de génération d'images et d'inférence de variables latentes. L'un des modèles typiques est le modèle Iterative Adversarial Inference (GibbsNet), qui apprend la distribution conjointe entre les données et leur variable latente. Nous présentons RGNet (Re-inference GibbsNet) qui introduit une chaîne de réinférence dans GibbsNet pour améliorer la qualité des échantillons générés et des variables latentes déduites. RGNet comprend les réseaux génératifs, d'inférence et discriminatifs. Un jeu contradictoire se déroule entre les réseaux génératifs et d'inférence et le réseau discriminatif. Le réseau discriminant est entraîné à distinguer (i) les paires conjointes inférence-latente/espace de données et les paires ré-inférence-latente/espace de données et (ii) les paires conjointes échantillonné-latent/espace de données généré. Nous montrons empiriquement que RGNet surpasse GibbsNet dans la qualité des variables latentes inférées et atteint des performances comparables sur les tâches de génération d'images et d'inpainting.
Zhihao LIU
Beijing Jiaotong University
Hui YIN
Beijing Jiaotong University
Hua HUANG
Beijing Jiaotong University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copier
Zhihao LIU, Hui YIN, Hua HUANG, "Iterative Adversarial Inference with Re-Inference Chain for Deep Graphical Models" in IEICE TRANSACTIONS on Information,
vol. E102-D, no. 8, pp. 1586-1589, August 2019, doi: 10.1587/transinf.2018EDL8256.
Abstract: Deep Graphical Model (DGM) based on Generative Adversarial Nets (GANs) has shown promise in image generation and latent variable inference. One of the typical models is the Iterative Adversarial Inference model (GibbsNet), which learns the joint distribution between the data and its latent variable. We present RGNet (Re-inference GibbsNet) which introduces a re-inference chain in GibbsNet to improve the quality of generated samples and inferred latent variables. RGNet consists of the generative, inference, and discriminative networks. An adversarial game is cast between the generative and inference networks and the discriminative network. The discriminative network is trained to distinguish between (i) the joint inference-latent/data-space pairs and re-inference-latent/data-space pairs and (ii) the joint sampled-latent/generated-data-space pairs. We show empirically that RGNet surpasses GibbsNet in the quality of inferred latent variables and achieves comparable performance on image generation and inpainting tasks.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2018EDL8256/_p
Copier
@ARTICLE{e102-d_8_1586,
author={Zhihao LIU, Hui YIN, Hua HUANG, },
journal={IEICE TRANSACTIONS on Information},
title={Iterative Adversarial Inference with Re-Inference Chain for Deep Graphical Models},
year={2019},
volume={E102-D},
number={8},
pages={1586-1589},
abstract={Deep Graphical Model (DGM) based on Generative Adversarial Nets (GANs) has shown promise in image generation and latent variable inference. One of the typical models is the Iterative Adversarial Inference model (GibbsNet), which learns the joint distribution between the data and its latent variable. We present RGNet (Re-inference GibbsNet) which introduces a re-inference chain in GibbsNet to improve the quality of generated samples and inferred latent variables. RGNet consists of the generative, inference, and discriminative networks. An adversarial game is cast between the generative and inference networks and the discriminative network. The discriminative network is trained to distinguish between (i) the joint inference-latent/data-space pairs and re-inference-latent/data-space pairs and (ii) the joint sampled-latent/generated-data-space pairs. We show empirically that RGNet surpasses GibbsNet in the quality of inferred latent variables and achieves comparable performance on image generation and inpainting tasks.},
keywords={},
doi={10.1587/transinf.2018EDL8256},
ISSN={1745-1361},
month={August},}
Copier
TY - JOUR
TI - Iterative Adversarial Inference with Re-Inference Chain for Deep Graphical Models
T2 - IEICE TRANSACTIONS on Information
SP - 1586
EP - 1589
AU - Zhihao LIU
AU - Hui YIN
AU - Hua HUANG
PY - 2019
DO - 10.1587/transinf.2018EDL8256
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E102-D
IS - 8
JA - IEICE TRANSACTIONS on Information
Y1 - August 2019
AB - Deep Graphical Model (DGM) based on Generative Adversarial Nets (GANs) has shown promise in image generation and latent variable inference. One of the typical models is the Iterative Adversarial Inference model (GibbsNet), which learns the joint distribution between the data and its latent variable. We present RGNet (Re-inference GibbsNet) which introduces a re-inference chain in GibbsNet to improve the quality of generated samples and inferred latent variables. RGNet consists of the generative, inference, and discriminative networks. An adversarial game is cast between the generative and inference networks and the discriminative network. The discriminative network is trained to distinguish between (i) the joint inference-latent/data-space pairs and re-inference-latent/data-space pairs and (ii) the joint sampled-latent/generated-data-space pairs. We show empirically that RGNet surpasses GibbsNet in the quality of inferred latent variables and achieves comparable performance on image generation and inpainting tasks.
ER -