paper-with-me

홈 › Papers

How to Construct Energy for Images? Denoising Autoencoder Can Be Energy Based Model

2023-03-05 · Weili Zeng

Energy-based models parameterize the unnormalized log-probability of data samples, but there is a lack of guidance on how to construct the "energy". In this paper, we propose a Denoising-EBM which decomposes the image energy into "semantic energy" and "texture energy". We define the "semantic energy" in the latent space of DAE to model the high-level representations, and define the pixel-level reconstruction error for denoising as "texture energy". Inspired by score-based model, our model utilizes multi-scale noisy samples for maximum-likelihood training and it outputs a vector instead of a scalar for exploring a larger set of functions during optimization. After training, the semantics are first synthesized by fast MCMC through "semantic energy", and then the pixel-level refinement of semantic image will be performed to generate perfect samples based on "texture energy". Ultimately, our model can outperform most EBMs in image generation. And we also demonstrate that Denoising-EBM has top performance among EBMs for out-of-distribution detection.

📄 PDF Abstract BibTeX arXiv:2303.03887

Code (0)

등록된 구현이 없습니다.

Tasks

DenoisingImage GenerationOut-of-Distribution Detection

Similar Papers 제목 키워드 기반

Deep Energy Estimator Networks

2018-05-21 · Saeed Saremi, Arash Mehrjou, Bernhard Schölkopf, Aapo Hyvärinen

Density estimation is a fundamental problem in statistical learning. This problem is especially challenging for complex high-dimensional data due to the curse of dimensionality. A promising solution to this problem is gi…

DenoisingDensity Estimation

Iterative energy-based projection on a normal data manifold for anomaly localization

2020-02-10 · ICLR 2020 1 · David Dehaene, Oriel Frigo, Sébastien Combrexelle, Pierre Eline

Autoencoder reconstructions are widely used for the task of unsupervised anomaly localization. Indeed, an autoencoder trained on normal data is expected to only be able to reconstruct normal features of the data, allowin…

Anomaly DetectionAnomaly LocalizationUnsupervised Anomaly Detection

Reconstruct & Crush Network

2017-12-01 · NeurIPS 2017 12 · Erinc Merdivan, Mohammad Reza Loghmani, Matthieu Geist

This article introduces an energy-based model that is adversarial regarding data: it minimizes the energy for a given data distribution (the positive samples) while maximizing the energy for another given data distributi…

Rate-Optimal Denoising with Deep Neural Networks

2018-05-22 · ICLR 2019 5 · Reinhard Heckel, Wen Huang, Paul Hand, Vladislav Voroninski

Deep neural networks provide state-of-the-art performance for image denoising, where the goal is to recover a near noise-free image from a noisy observation. The underlying principle is that neural networks trained on la…

DenoisingImage Denoising

Attention in a family of Boltzmann machines emerging from modern Hopfield networks

2022-12-09 · Toshihiro Ota, Ryo Karakida

Hopfield networks and Boltzmann machines (BMs) are fundamental energy-based neural network models. Recent studies on modern Hopfield networks have broaden the class of energy functions and led to a unified perspective on…

Denoising