Optimal Mass Transport: Signal processing and machine-learning applications

2017/07/01 by Soheil Kolouri, Se Rim Park, Matthew Thorpe +3 · 16 citations
Computer Science · Physics and Astronomy · #Medical Image Segmentation Techniques #Topological and Geometric Data Analysis #Statistical Mechanics and Entropy

paper · doi:10.1109/msp.2017.2695801

Abstract

Transport-based techniques for signal and data analysis have received increased attention recently. Given their ability to provide accurate generative models for signal intensities and other data distributions, they have been used in a variety of applications including content-based retrieval, cancer detection, image super-resolution, and statistical machine learning, to name a few, and shown to produce state of the art results in several applications. Moreover, the geometric characteristics of transport-related metrics have inspired new kinds of algorithms for interpreting the meaning of data distributions. Here we provide a practical overview of the mathematical underpinnings of mass transport-related methods, including numerical implementation, as well as a review, with demonstrations, of several applications. Software accompanying this tutorial is available at [43].

Citations

• Learning with a Wasserstein Loss
• Wasserstein Training of Boltzmann Machines
• A computational fluid mechanics solution to the Monge-Kantorovich mass transfer problem
• The geometry of optimal transportation
• Polar factorization and monotone rearrangement of vector‐valued functions
• The Earth Mover's Distance as a Metric for Image Retrieval

Cited by

• Scalable Dynamic Optimal Transport via Distributed Linearized ADMM
• Power spectrum signatures of graphs
• Enpowering Your Pansharpening Models with Generalizability: Unified Distribution is All You Need
• Biology-driven assessment of deep learning super-resolution imaging of the porosity network in dentin
• InforME: Improving Informativeness of Abstractive Text Summarization With Informative Attention Guided by Named Entity Salience
• Well-Posedness and Efficient Algorithms for Inverse Optimal Transport with Bregman Regularization
• Quantifying functionally equivalent species and ecological network dissimilarity with optimal transport distances
• Wasserstein Distributionally Robust Optimization: Theory and\n Applications in Machine Learning
• Efficient Sliced Wasserstein Distance Computation via Adaptive Bayesian Optimization
• E-ROBOT: a dimension-free method for robust statistics and machine learning via Schrödinger bridge
• New Metrics Between Rational Spectra and their Connection to Optimal Transport
• Cooperative System Identification via Correctional Learning
• Inverse harmonic clustering for multi-pitch estimation: an optimal transport approach
• Statistical and Topological Properties of Gaussian Smoothed Sliced Probability Divergences
• Learning Adversarially Robust Representations via Worst-Case Mutual Information Maximization
• Optimal Transport Relaxations with Application to Wasserstein GANs

Related

• Transport-based analysis, modeling, and learning from signal and data distributions
• A Survey on Algorithmic Developments in Optimal Transport Problem with Applications
• Data representation with optimal transport
• Statistical Inference for Optimal Transport Maps: Recent Advances and Perspectives
• A Survey on Optimal Transport for Machine Learning: Theory and Applications