Adaptive Multi-stage Density Ratio Estimation for Learning Latent Space Energy-based Model

被引:0
|
作者
Xiao, Zhisheng [1 ]
Han, Tian [2 ]
机构
[1] Univ Chicago, Chicago, IL 60637 USA
[2] Stevens Inst Technol, Hoboken, NJ 07030 USA
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
This paper studies the fundamental problem of learning energy-based model (EBM) in the latent space of the generator model. Learning such prior model typically requires running costly Markov Chain Monte Carlo (MCMC). Instead, we propose to use noise contrastive estimation (NCE) to discriminatively learn the EBM through density ratio estimation between the latent prior density and latent posterior density. However, the NCE typically fails to accurately estimate such density ratio given large gap between two densities. To effectively tackle this issue and learn more expressive prior models, we develop the adaptive multi-stage density ratio estimation which breaks the estimation into multiple stages and learn different stages of density ratio sequentially and adaptively. The latent prior model can be gradually learned using ratio estimated in previous stage so that the final latent space EBM prior can be naturally formed by product of ratios in different stages. The proposed method enables informative and much sharper prior than existing baselines, and can be trained efficiently. Our experiments demonstrate strong performances in image generation and reconstruction as well as anomaly detection.
引用
收藏
页数:12
相关论文
共 50 条
  • [31] Adaptive Cyber Defense Against Multi-Stage Attacks Using Learning-Based POMDP
    Hu, Zhisheng
    Zhu, Minghui
    Liu, Peng
    ACM TRANSACTIONS ON PRIVACY AND SECURITY, 2021, 24 (01)
  • [32] Molecule Design by Latent Space Energy-Based Modeling and Gradual Distribution Shifting
    Kong, Deqian
    Pang, Bo
    Han, Tian
    Wu, Ying Nian
    UNCERTAINTY IN ARTIFICIAL INTELLIGENCE, 2023, 216 : 1109 - 1120
  • [33] ACCURATE HEAD POSE ESTIMATION BASED ON MULTI-STAGE REGRESSION
    Liu, Yinchuan
    Gong, Yufei
    Lu, Zheng
    Zhang, Xuetao
    2022 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING, ICIP, 2022, : 1326 - 1330
  • [34] Adaptive multi-stage 2D image motion field estimation
    Neumann, U
    You, SY
    APPLICATIONS OF DIGITAL IMAGE PROCESSING XXI, 1998, 3460 : 116 - 123
  • [35] A multi-stage machine learning model for diagnosis of esophageal manometry
    Kou, Wenjun
    Carlson, Dustin A.
    Baumann, Alexandra J.
    Donnan, Erica N.
    Schauer, Jacob M.
    Etemadi, Mozziyar
    Pandolfino, John E.
    ARTIFICIAL INTELLIGENCE IN MEDICINE, 2022, 124
  • [36] Model-free Reinforcement Learning based Multi-stage Smart Noise Jamming
    Wang, Yuanhang
    Zhang, Tianxian
    Xu, Longxiao
    Tian, Tuanwei
    Kong, Lingjiang
    Yang, Xiaobo
    2019 IEEE RADAR CONFERENCE (RADARCONF), 2019,
  • [37] A Multi-Stage Acoustic Echo Cancellation Model Based on Adaptive Filters and Deep Neural Networks
    Xu, Shiyun
    He, Changjun
    Yan, Bosong
    Wang, Mingjiang
    ELECTRONICS, 2023, 12 (15)
  • [38] Parallel multi-stage preconditioners with adaptive setup for the black oil model
    Zhao, Li
    Feng, Chunsheng
    Zhang, Chen-Song
    Shu, Shi
    COMPUTERS & GEOSCIENCES, 2022, 168
  • [39] Energy-Based Geometric Multi-model Fitting
    Isack, Hossam
    Boykov, Yuri
    INTERNATIONAL JOURNAL OF COMPUTER VISION, 2012, 97 (02) : 123 - 147
  • [40] Sentence Simplification Based on Multi-Stage Encoder Model
    Zhang, Lemin
    Deng, Huifang
    IEEE ACCESS, 2019, 7 : 174248 - 174256