A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Filters
Learning Energy-Based Models by Diffusion Recovery Likelihood
[article]
2020
Preserved Fulltext
Inspired by recent progress on diffusion probabilistic models, we present a diffusion recovery likelihood method to tractably learn and sample from a sequence of EBMs trained on increasingly noisy versions ...
While energy-based models (EBMs) exhibit a number of desirable properties, training and sampling on high-dimensional datasets remains challenging. ...
B EXPERIMENTAL DETAILS Model architecture. Our network structure is based on Wide ResNet (Zagoruyko & Komodakis, 2016) . Table 5 lists the detailed network structures of various resolutions. ...
doi:10.48550/arxiv.2012.08125
fatcat:q346wymmnnektfm4dhugblfehi
EBM Life Cycle: MCMC Strategies for Synthesis, Defense, and Density Modeling
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
This work presents strategies to learn an Energy-Based Model (EBM) according to the desired length of its MCMC sampling trajectories. ...
To achieve these outcomes, we introduce three novel methods of MCMC initialization for negative samples used in Maximum Likelihood (ML) learning. ...
This approach is taken by Energy-Based Models (EBMs) [1] , normalizing flows [2] , score-based models [3, 4] , and auto-regressive models [5] , as well as by Variational Auto-encoders (VAEs) [6] ...
arXiv:2205.12243v1
fatcat:sbntc3otc5gi5dvjan4z24mt2m
A New Search for Neutrino Point Sources with IceCube
[article]
2021
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Other Versions
We replaced the usual Gaussian approximations of IceCube's point spread function with precise numerical representations, obtained from simulations, and combined them with new machine learning-based estimates ...
While eight years have passed since IceCube discovered a diffuse flux of high-energy astrophysical neutrinos, the sources of the vast majority of these neutrinos remain unknown. ...
and truncated energy, a traditional likelihood-based algorithm (right). ...
arXiv:2107.08700v2
fatcat:bsfi3s7mjvgilfoujtwjrwje4u
Energy-Based Generative Cooperative Saliency Prediction
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Other Versions
Specifically, we propose a generative cooperative saliency prediction framework, where a conditional latent variable model (LVM) and a conditional energy-based model (EBM) are jointly trained to predict ...
In this paper, to model the uncertainty of visual saliency, we study the saliency prediction problem from the perspective of generative models by learning a conditional probability distribution over the ...
The maximum likelihood learning of the energy-based model typically requires iterative MCMC sampling, which is computationally challenging. ...
arXiv:2106.13389v2
fatcat:lwvah5fvrzbxxfyfdphujxjjai
Learning Energy-based Spatial-Temporal Generative ConvNets for Dynamic Patterns
[article]
2019
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
We show that an energy-based spatial-temporal generative ConvNet can be used to model and synthesize dynamic patterns. ...
The model can be learned from the training video sequences by an "analysis by synthesis" learning algorithm that iterates the following two steps. ...
The work is supported by ONR MURI N00014-16-1-2007, and DARPA ARO W911NF-16-1-0579. ...
arXiv:1909.11975v1
fatcat:celbwe5qfjabjhyoiagr464fxi
Guest Editorial: Special Issue in Memory of Mila Nikolova
2020
Journal of Mathematical Imaging and Vision
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
The authors of Stable backward diffusion models that minimise convex energies (L. Bergerhoff, M. Cárdenas, J. Weickert, M. ...
The challenging problem of fine structure detection with applications to bituminous surfacing crack recovery is examined in A nonlocal Laplacian-based model for bituminous surfacing crack recovery and ...
doi:10.1007/s10851-020-00981-6
fatcat:c5xojwqquvce5lyhujgcsd5u3e
Recovery of relative depth from a single observation using an uncalibrated (real-aperture) camera
2008
2008 IEEE Conference on Computer Vision and Pattern Recognition
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
Because of ill-posedness we propose a graph-cuts based method for inferring the depth in the scene using the amount of diffusion as a data likelihood and a smoothness condition on the depth in the scene ...
We stabilize the reverse heat equation by considering the gradient degeneration as an effective stopping criterion. The amount of (inverse) diffusion is actually a measure of relative depth. ...
A successful approach for single image based structure recovery has been that by Criminisi et al. [5] . ...
doi:10.1109/cvpr.2008.4587779
dblp:conf/cvpr/NamboodiriC08
fatcat:jtripwnefnbsxd6l3wzastx36q
Latent Diffusion Energy-Based Model for Interpretable Text Modeling
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Other Versions
Inspired by the recent efforts that leverage diffusion recovery likelihood learning as a cure for the sampling issue, we introduce a novel symbiosis between the diffusion models and latent space EBMs in ...
a variational learning framework, coined as the latent diffusion energy-based model. ...
Wu was supported by NSF DMS-2015577. We would like to thank the anonymous reviewers for their constructive comments. ...
arXiv:2206.05895v3
fatcat:kktenwa5qbevtbpuzcik6rrx4a
Neural network analysis of X-ray polarimeter data
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
and model uncertainties. ...
Deep neural network based classifiers can be used to select against these events to improve energy resolution and polarization sensitivity. ...
Photon energy. Energy recovery can be adversely affected by tail events. ...
arXiv:2206.10537v1
fatcat:egtxzwoklvesjgy5u7va2pj2s4
Introducing Dynamic Prior Knowledge to Partially-Blurred Image Restoration
[chapter]
2006
Lecture Notes in Computer Science
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
To estimate the point spread function (PSF), a parametric model space is introduced to reduce the searching uncertainty for PSF model selection. ...
Maximum a posteriori estimation of parameters in Bayesian regularization is equal to minimizing energy of a dataset for a given number of classes. ...
Based on the model space, an unsupervised self-initializing PSF learning term can learn a PSF parametric model according to the following energy functional, F (ĥ|f , g) = 1 2 Ω (g −ĥ * f ) 2 dA + β Ω | ...
doi:10.1007/11861898_12
fatcat:lyb7ptyzdbbvvdm2b7h2xtxgqq
2021 Index IEEE Transactions on Signal Processing Vol. 69
2021
IEEE Transactions on Signal Processing
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
The primary entry includes the coauthors' names, the title of the paper or other item, and its location, specified by the publication abbreviation, year, month, and inclusive pagination. ...
., +, TSP 2021 5830-5845 Location based services Diffusion SLAM: Localizing Diffusion Sources From Samples Taken by Location-Unaware Mobile Sensors. ...
Imtiaz, H., +, TSP 2021 6355-6370 Diffusion Diffusion SLAM: Localizing Diffusion Sources From Samples Taken by Location-Unaware Mobile Sensors. ...
doi:10.1109/tsp.2022.3162899
fatcat:kcubj566gzb4zkj7xb5r5we3ri
Minimum Probability Flow Learning
[article]
2011
arXiv
pre-print
Preserved Fulltext
The Internet Archive has a preservation copy of this work in our general collections.
The file type is application/pdf.
Other Versions
In the Ising model case, current state of the art techniques are outperformed by at least an order of magnitude in learning time, with lower error in recovered coupling parameters. ...
Score matching, minimum velocity learning, and certain forms of contrastive divergence are shown to be special cases of this learning technique. ...
Schapire and William Bialek for use of their Ising model coupling parameters; Jonathon Shlens for useful discussion and ground truth for his Ising model convergence times; Bruno Olshausen, Anthony Bell ...
arXiv:0906.4779v4
fatcat:vbhj2juwcfgmvldujod7sklzu4
Table of Contents
2021
IEEE Transactions on Signal Processing
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Blu Diffusion SLAM: Localizing Diffusion Sources From Samples Taken by Location-Unaware Mobile Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...
Trung Marginal Likelihood Maximization Based Fast Array Manifold Matrix Learning for Direction of Arrival Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...
doi:10.1109/tsp.2021.3136800
fatcat:zhf46mb3rbdlnnh3u2xizgxof4
Prediction-Centric Learning of Independent Cascade Dynamics from Partial Observations
[article]
2021
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Other Versions
We address the problem of learning of a spreading model such that the predictions generated from this model are accurate and could be subsequently used for the optimization, and control of diffusion dynamics ...
We introduce a computationally efficient algorithm, based on a scalable dynamic message-passing approach, which is able to learn parameters of the effective spreading model given only limited information ...
In the recent years, the problem of diffusion model recovery from full observations has been extensively addressed through a series of works focusing on heuristic and exact algorithms, largely based on ...
arXiv:2007.06557v3
fatcat:g3xamqwhujcotg4yihsiugkvx4
Learning Energy-Based Models With Adversarial Training
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Other Versions
We study a new approach to learning energy-based models (EBMs) based on adversarial training (AT). ...
We show that (binary) AT learns a special kind of energy function that models the support of the data distribution, and the learning process is closely related to MCMC-based maximum likelihood learning ...
.: Learning latent space energy-based prior model. ...
arXiv:2012.06568v3
fatcat:ba4ya6kxsjd2ti2bw4borauc34
« Previous
Showing results 1 — 15 out of 15,303 results