Welcome to ProxImaL — ProxImaL 0.1.3 documentation

Time 2022-10-20 11:29:03

Web Name: Welcome to ProxImaL — ProxImaL 0.1.3 documentation

WebSite: http://www.proximal-lang.org

ID:359744

Keywords:

ProxImaL,to,Welcome,documentation

Description:

Welcome to ProxImaL¶

ProxImaL is a Python-embedded modeling language for image optimization problems.It allows you to express your problem in a natural way that follows the math,and automatically determines an efficient method for solving the problem.ProxImaL makes it easy to experiment with many different priors and other problem reformulations,without worrying about the details of how the problem is solved.

For example, the following code denoises an image using simple sparse gradient and nonnegativity priors:

from proximal import *import numpy as npimport scipy.miscimport matplotlib.pyplot as plt# Generate data.I = scipy.misc.ascent()np.random.seed(1)b = I + 10*np.random.randn(*I.shape)# Construct and solve problem.x = Variable( I.shape )prob = Problem(sum_squares(x - b/255) + .1*norm1( grad(x) ) + nonneg(x))prob.solve()# Plot the original, noisy, and denoised images.plt.figure(figsize=(15,8))plt.subplot(131)plt.gray()plt.imshow(I)plt.title('Original image')plt.subplot(132)plt.gray()plt.imshow(b)plt.title('Noisy image')plt.subplot(133)plt.gray()plt.imshow(x.value*255) # x.value is the optimal value of x.plt.title('Denoising results')plt.show()

The example above uses simple, well-known priors. Much better results can be obtained using the more sophisticated priors provided in ProxImaL.

ProxImaL was designed and implemented by Felix Heide and Steven Diamond.See the accompanying paper for a full discussion of ProxImaL’s design and examples of state-of-the-art results obtained with ProxImaL.

ProxImaL

Navigation

Install GuideTutorialPublications

Quick search

Enter search terms or a module, class or function name.

©2016, Felix Heide and Steven Diamond. | Powered by Sphinx 1.3.5 & Alabaster 0.7.9 | Page source

TAGS:ProxImaL to Welcome documentation

<<< Thank you for your visit >>>

Hot Websites