A metric for distributions with applications to image databases

By Yossi Rubner, Carlo Tomasi, Leonidas J. Guibas. 1998

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Multimodal Explanations: Justifying Decisions and Pointing to the Evidence (2018, methodology): “We use the code from [25] to compute EMD.”
Auditing black-box models for indirect influence (2016, methodology): “distribution given by A, then O is maximally obscured, but W also minimally changes, in that A also minimizes the function ∑ x∈O d(Wx, A) where d(·, ·) was the earthmover distance [23] between the distributions using `2 as the base metric.”
DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation (2019, background): “Earth Mover’s distance [45], also known as the Wasserstein distance, is another popular metric for measuring the difference between two discrete distributions.”
Evaluation of Text Generation: A Survey (2020, background): “Word Mover’s Distance (WMD) Earth mover’s distance (EMD), also known as the Wasserstein metric (Rubner et al., 1998), is a measure of the distance between two probability distributions.”
Optimal Transport for Domain Adaptation (2014, neutral): “The Wasserstein distance is also known as the Earth Mover Distance in the computer vision community [40] and it defines a metric over the space of integrable squared probability measures.”
Attentive Explanations: Justifying Decisions and Pointing to the Evidence (2016, methodology): “For visual pointing task, we use Earth Mover’s Distance (EMD) [30], which measures the distance between two probability distributions over a region, and rank correlation, which was used in [11], as our evaluation metrics.”
Supervised Learning of Semantic Classes for Image Annotation and Retrieval (2007, methodology): “While early retrieval architectures were based on the query-by-example paradigm [7], [17], [18], [19], [24], [25], [26], [30], [32], [35], [37], [39], [45], which formulates image retrieval as the search for the best database match to a user-provided query image, it was quickly realized that the des”
A survey of network anomaly detection techniques (2016, methodology): “EMD considers cross-bin matching and provides a more accurate evaluation on the dissimilarity between distributions than some other well-known dissimilarity measures.”
On measuring the distance between histograms (2002, methodology): “The EMD considers more general ground distances that it works for multivariate histograms as well as weighted clusters [14].”
The Earth Mover's distance is the Mallows distance: some insights from statistics (2001, methodology): “Extensive empirical evidence on the usefulness of the EMD for texture analysis and image retrieval has already been published [11, 12, 9]; therefore we chose not to conduct large-scale experiments.”
Sliced and Radon Wasserstein Barycenters of Measures (2014, unknown): “The Wasserstein distance leads to state-of-the-art results for several image retrieval problems, see for instance [30] for an early work on this topic.”
Model-Based Multiple Instance Learning (2017, methodology): “Existing classifiers in the Bag-Space paradigm are based on distances between point patterns, such as Hausdorff [34], [35], Chamfer [36], Earth Mover’s [37], [38].”
Supervised Word Mover's Distance (2016, background): “Cuturi & Avis [7] was the first to consider learning the ground distance in the Earth Mover’s Distance (EMD).”
Visualizing structure and transitions in high-dimensional biological data (2019, background): “The EMD is measure of dissimilarity between two probability distributions that is particularly popular in computer vision 73 .”
Multitask Bregman clustering (2010, background): “Actually, earth mover distance (EMD) [37] is a convenient choice for this problem, which is widely used in music analysis [38] and computer vision [37].”
Do Language Embeddings capture Scales? (2020, background): “3 Earth Mover’s Distance (EMD) Also known as the Wasserstein distance (Rubner et al., 1998).”