Scalable Optimal Transport

A curated list of research works and resources on optimal transport in machine learning. Related Paper.

@article{khamis24OT,
      title={Scalable Optimal Transport Methods in Machine Learning: A Contemporary Survey}, 
      author={Abdelwahed Khamis and Russell Tsuchida and Mohamed Tarek and Vivien Rolland and Lars Petersson},
      year={2024},
      journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
      publisher={IEEE}
}

Background

2021 - SIAM Stochastic control liaisons: Richard sinkhorn meets gaspard monge on a schrodinger bridge
2019 - Computational optimal transport: With applications to data science
2015 - Optimal transport for applied mathematicians: calculus of variations, PDEs, and modeling
2009 - Optimal transport: old and new
2093 - Topics in optimal transportation
2006 - On a problem of monge

Regularized OT

2023 - CoRL Watch and match: Supercharging imitation with regularized optimal transport Code
2022 - IG Entropy-regularized 2-wasserstein distance between gaussian measures
2021 - ICDM A regularized wasserstein framework for graph kernels
2020 - TMLR MMD-regularized Unbalanced Optimal Transport Code
2020 - ICML Regularized optimal transport is ground cost adversarial
2020 - SIAM Empirical regularized optimal transport: Statistical theory and applications
2019 - Quantum entropic regularization of matrix-valued optimal transport
2019 - SIAM Stabilized sparse scaling algorithms for entropy regularized transport problems
2018 - SIAM Semidual regularized optimal transport
2018 - AISTATS Smooth and sparse optimal transport
2014 - NeurIPS Optimal transport with laplacian regularization
2014 - SIAM Regularized discrete optimal transport Code

Sinkhorn

2022 - CVPR A unified framework for implicit sinkhorn differentiation
2021 - ICML Low-rank sinkhorn factorization
2021 - A note on overrelaxation in the sinkhorn algorithm
2020 - NeurIPS Faster wasserstein distance estimation with the sinkhorn divergence Code
2020 - ICML On unbalanced optimal transport: An analysis of sinkhorn algorithm Code
2020 - NeurIPS Linear time sinkhorn divergences using positive features
2019 - NeurIPS Massively scalable sinkhorn distances via the nyström method
2019 - AISTATS Interpolating between optimal transport and mmd using sinkhorn divergences
2019 - arXiv Sinkhorn divergences for unbalanced optimal transport
2019 - AISTATS Sample complexity of sinkhorn divergences
2019 - ICML Wasserstein adversarial examples via projected sinkhorn iterations
2019 - NeurIPS Differentiable Ranking and Sorting using Optimal Transport
2018 - ICML Computational optimal transport: Complexity by accelerated gradient descent is better than by sinkhorn’s algorithm
2018 - ICLR Learning latent permutations with gumbel-sinkhorn networks Code
2017 - NeurIPS Overrelaxed sinkhorn-knopp algorithm for regularized optimal transport
2013 - NeurIPS Sinkhorn Distances: Lightspeed Computation of Optimal Transport Code
2011 - Ranking via sinkhorn propagation
2008 - SIAM The sinkhorn–knopp algorithm: convergence and applications

Unbalanced and Partial OT

2022 - arXiv Unbalanced optimal transport, from theory to numerics
2021 - ICPR Unbalanced optimal transport in multi-camera tracking applications
2021 - AAAI Learning to count via unbalanced optimal transport
2020 - NeurIPS Partial optimal tranport with applications on positive-unlabeled learning
2020 - NeurIPS Entropic optimal transport between unbalanced gaussian measures has a closed form
2018 - Unbalanced optimal transport: Dynamic and kantorovich formulations

Structured OT

2024 - ICML Submodular Framework for Structured-Sparse Optimal Transport Code
2023 - ICLR Sparsity-Constrained Optimal Transport
2019 - NeurIPS Hierarchical optimal transport for document representation Code
2018 - AISTATS Structured optimal transport

Dynamic OT & Schrodinger Bridge

2023 - arXiv Transport with support: Data-conditional diffusion bridges

Multi-Marginal & Wasserstein Barycenter

2021 - SIAM Multimarginal optimal transport with a tree-structured cost and the schrodinger bridge problem
2020 - ICML Debiased sinkhorn barycenters
2020 - NeurIPS Continuous regularized wasserstein barycenters Code
2020 - Multi-marginal optimal transport using partial information with applications in robust localization and sensor fusion
2016 - ACM ToG Wasserstein barycentric coordinates: Histogram regression using optimal transport
2015 - Sliced and radon wasserstein barycenters of measures Code
2011 - SIAM Barycenters in the wasserstein space

Tree Wasserstein

2023 - TMLR Approximating 1-wasserstein distance with trees
2022 - AISTATS Fixed support tree-sliced wasserstein barycenter Code
2021 - ICML Supervised tree-wasserstein distance
2020 - NeurIPS Fast unbalanced optimal transport on a tree Code
2019 - NeurIPS Tree-sliced variants of wasserstein distances Code

Gromov Wasserstein

2022 - A brief survey on computational gromov-wasserstein distance
2022 - arXiv Gromov-wasserstein autoencoders Code
2022 - ICML Entropic gromov-wasserstein between gaussian distributions
2021 - AISTATS Aligning time series on incomparable spaces
2021 - ECML PKDD Quantized gromov-wasserstein Code
2020 - ICML Gromov-Wasserstein Optimal Transport to Align Single-Cell Multi-Omics Data
2020 - A contribution to optimal transport on incomparable spaces
2019 - NeurIPS Asymptotic guarantees for learning generative models with the sliced-wasserstein distance
2019 - CVPR Max-sliced wasserstein distance and its use for gans
2018 - EMNLP Gromov-wasserstein alignment of word embedding spaces Code
2016 - ICML Gromov-wasserstein averaging of kernel and distance matrices Code
2011 - Gromov–wasserstein distances and the metric approach to object matching

Sliced Wasserstein

2021 - NeurIPS Pooling by sliced-wasserstein embedding Code
2021 - CVPR A sliced wasserstein loss for neural texture synthesis Code
2021 - ICML Differentially private sliced wasserstein distance
2020 - NeurIPS Statistical and topological properties of sliced probability divergences
2019 - NeurIPS Sliced gromov-wasserstein Code
2019 - NeurIPS Generalized sliced wasserstein distances
2016 - CVPR Sliced wasserstein kernels for probability distributions

Other formulations

2023 - ECML PKDD Feature-robust optimal transport for high-dimensional data
2022 - ICML Order constraints in optimal transport Code
2021 - ICML Outlier-robust optimal transport
2020 - AISTATS Unsupervised hierarchy matching with optimal transport over hyperbolic spaces
2020 - ICML A swiss army knife for minimax optimal transport Code
2020 - NeurIPS Co-optimal transport Code
2017 - CVPR Order-preserving wasserstein distance for sequence matching

Theoretical and Complexity

2022 - Estimation of wasserstein distances in the spiked transport model
2022 - NeurIPS Asymptotics of smoothed wasserstein distances in the small noise regime
2021 - NeurIPS Rates of estimation of optimal transport maps using plug-in estimators via barycentric projections
2021 - arXiv Plugin estimation of smooth optimal transport maps
2021 - arXiv A short proof on the rate of convergence of the empirical measure for the wasserstein distance
2021 - NeurIPS Averaging on the bures-wasserstein manifold: dimension-free convergence of gradient descent
2021 - NeurIPS Dimensionality reduction for wasserstein barycenter
2020 - Convergence and concentration of empirical measures under wasserstein distance in unbounded functional spaces
2020 - arXiv A study of performance of optimal transport Code
2020 - arXiv The statistical effect of entropic regularization in optimal transportation
2019 - Sharp asymptotic and finite-sample rates of convergence of empirical measures in wasserstein distance
2019 - arXiv Strong equivalence between metrics of wasserstein type
2018 - Optimal entropy-transport problems and a new hellinger–kantorovich distance between positive measures
2016 - SIAM A smoothed dual approach for variational wasserstein problems Code

Scaling OT

Measures simplification

2022 - AMTA Quantized wasserstein procrustes alignment of word embedding spaces
2022 - WIRE CS Projection-based techniques for high-dimensional optimal transport problems
2021 - UAI Improving approximate optimal transport distances using quantization
2021 - MDPI-ASubspace detours meet gromov–wasserstein
2021 - ICML Projection robust wasserstein barycenters
2020 - AISTATS Gaussian-smoothed optimal transport: Metric structure and statistical efficiency
2019 - ICML Subspace robust wasserstein distances Code
2019 - NeurIPS Subspace detours: Building transport plans that are optimal on subspace projections
2019 - NeurIPS Large-scale optimal transport map estimation using projection pursuit Code

Cost and Kernel Approximation

2021 - ICML Scalable optimal transport in high dimensions for graph distances, embedding alignment, and more Code
2019 - SIAM Stabilized sparse scaling algorithms for entropy regularized transport problems

Plan Structuring

2022 - NeurIPS Low-rank optimal transport: Approximation, statistics and debiasing
2022 - ICML Linear-time gromov wasserstein distances using low rank couplings and costs
2021 - arXiv Approximating optimal transport via low-rank and sparse factorization
2021 - Making transport more robust and interpretable by moving data through a small number of anchor points
2019 - AISTATS Statistical optimal transport via factored couplings

Computation Partitioning

2022 - arXiv Budget-constrained bounds for mini-batch estimation of optimal transport
2022 - CVPR Computing wasserstein-p distance between images with linear cost
2021 - Deep learning and optimal transport: learning from one another
2020 - arXiv Mrec: a fast and versatile framework for aligning and matching point clouds with applications to single cell molecular data
2020 - AISTATS Learning with minibatch wasserstein: asymptotic and gradient properties Code
2019 - JMLR Optimal transport: Fast probabilistic approximation with exact solvers.
2019 - NeurIPS Scalable gromov-wasserstein learning for graph partitioning and matching Code
2017 - JMLR Multiscale strategies for computing optimal transport Code
2011 - CGF A multiscale approach to optimal transport

Solver Acceleration

2023 - AISTATS Rethinking initialization of the sinkhorn algorithm
2022 - JMIV Learning to generate wasserstein barycenters Code
2022 - NeurIPS Supervised training of conditional monge maps
2022 - NeurIPS Wasserstein iterative networks for barycenter estimation Code
2022 - Kantorovich strikes back! wasserstein gans are not optimal transport?
2022 - AAAI Efficient optimal transport algorithm by accelerated gradient descent
2022 - AAAI Exploiting problem structure in deep declarative networks: Two case studies
2022 - arXiv Meta optimal transport Code
2021 - arXiv Wasserstein gans work because they fail (to approximate the wasserstein distance)
2021 - ICML Scalable computations of wasserstein barycenter via input convex neural networks
2021 - NeurIPS Do neural optimal transport solvers work? a continuous wasserstein-2 benchmark
2020 - ECAI Speeding up word mover’s distance and its variants via properties of distances between embeddings
2020 - UAI A fast proximal point method for computing exact wasserstein distance
2020 - ICML Optimal transport mapping via input convex neural networks
2018 - ICLR Large-scale optimal transport and mapping estimation Code
2017 - Linear-complexity relaxed word mover's distance with gpu acceleration
2016 - NeurIPS Stochastic optimization for large-scale optimal transport Code

Applications

Generative Models

2022 - NeurIPS Score-based Generative Modeling Secretly Minimizes the Wasserstein Distance Code
2021 - NeurIPS Maximum likelihood training of score-based diffusion models Code
2021 - NeurIPS Diffusion schrödinger bridge with applications to score-based generative modeling
2021 - ICLR Distributional sliced-wasserstein and applications to generative modeling
2021 - AAAI Towards generalized implementation of wasserstein distance in gans
2020 - NeurIPS Asymptotic guarantees for generative modeling based on the smooth wasserstein distance
2020 - ICLR Wasserstein-2 generative networks Code
2019 - ICML Learning generative models across incomparable spaces Code
2018 - AISTATS Learning generative models with sinkhorn divergences
2017 - ICML Wasserstein generative adversarial networks Code
2017 - From optimal transport to generative modeling: the vegan cookbook
2017 - NeurIPS Improved training of wasserstein gans Code

Domain Adaptation

2022 - ML Hierarchical optimal transport for unsupervised domain adaptation Code
2022 - ICLR Cross-domain imitation learning via optimal transport Code
2022 - IEEE TIP Few-shot domain adaptation via mixup optimal transport
2021 - ICML Unbalanced minibatch optimal transport; applications to domain adaptation Code
2021 - CVPR Wasserstein contrastive representation distillation
2021 - NeurIPS Lifelong domain adaptation via consolidated internal distribution
2021 - CVPR OTCE: A transferability metric for cross-domain cross-task representations Code
2021 - ICCV The right to talk: An audio-visual transformer approach Code
2021 - WACV Zero-shot recognition via optimal transport
2020 - CVPR Deepemd: Few-shot image classification with differentiable earth mover's distance and structured classifiers Code
2020 - ICML Margin-aware adversarial domain adaptation with optimal transport Code
2020 - ECCV Learning to generate novel domains for domain generalization Code
2020 - IJCAI Joint partial optimal transport for open set domain adaptation.
2020 - BMVC Weakly supervised cross-domain alignment with optimal transport
2020 - ICCV Transporting labels via hierarchical optimal transport for semi-supervised learning
2020 - NeurIPS Geometric dataset distances via optimal transport Code
2019 - CVPR Sliced wasserstein discrepancy for unsupervised domain adaptation Code
2019 - AISTATS Optimal transport for multi-source domain adaptation under target shift Code
2019 - NeurIPS Hierarchical optimal transport for multimodal distribution alignment
2018 - AAAI Wasserstein distance guided representation learning for domain adaptation
2018 - ICCV Deepjdot: Deep joint distribution optimal transport for unsupervised domain adaptation
2017 - ECML PKDD Theoretical analysis of domain adaptation with optimal transport
2017 - NeurIPS Joint distribution optimal transportation for domain adaptation Code
2016 - TPAMI Optimal transport for domain adaptation Code

Other Applications

2022 - CVPR Motion-modulated temporal fragment alignment network for few-shot action recognition Action Recognition
2022 - AISTATS Sinkformers: Transformers with doubly stochastic attention Transformers
2022 - AISTATS Proximal optimal transport modeling of population dynamics Code Modeling Dynamics
2022 - NeurIPS Optimal transport of classifiers to fairness Code Fairness in ML
2022 - CVPR Unsupervised action segmentation by joint representation learning and online clustering Action Segmentation
2021 - CVPR Ota: Optimal transport assignment for object detection Code Object Detection
2021 - ICCV Point-set distances for learning representations of 3d point clouds Code Point Cloud
2021 - CVPR A generalized loss function for crowd counting and localization Code Crowd Counting
2021 - WACV Augmented self-labeling for source-free unsupervised domain adaptation Self Labelling
2021 - NeurIPS Measuring generalization with optimal transport Code Generalization in ML
2021 - ICLR Convex potential flows: Universal probability distributions with optimal transport and convex optimization Normalizing Flow
2020 - NeurIPS Model fusion via optimal transport Code Model Fusion
2020 - ECCV Solving the blind perspective-n-point problem end-to-end with robust differentiable geometric optimization Code PnP problem
2019 - ICLR Self-labelling via simultaneous clustering and representation learning Code Self Labelling
2019 - ICML Obtaining fairness using optimal transport theory Code Fairness in ML
2019 - ICLR Learning embeddings into entropic wasserstein spaces Code Embedding
2015 - ICML From word embeddings to document distances Code Document Matching
2018 - TPAMI Visual permutation learning Code Permutation Learning
Correspondance & Matching:
- 2021 - CVPR Loftr: Detector-free local feature matching with transformers Code
- 2020 - CVPR Superglue: Learning feature matching with graph neural networks Code
- 2020 - arXiv Learning 3d-3d correspondences for one-shot partial-to-partial registration
Graphs:
- 2022 - ICML Learning to predict graphs with fused gromov-wasserstein barycenters
- 2020 - NeurIPS Copt: Coordinated optimal transport on graphs Code
- 2020 - ICLR Deep graph matching consensus Code
- 2019 - ICML Optimal transport for structured data with application on graphs Code
- 2019 - ICML Gromov-wasserstein learning for graph matching and node embedding Code

Software and Tools

2022 - arXiv Optimal transport tools (ott): A jax toolbox for all things wasserstein Code
2021 - JMLR Pot: Python optimal transport Code
2020 - NeurIPS Fast geometric learning with symbolic matrices Code

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Files

README.md

README.md

Scalable Optimal Transport

Contents

Background

Regularized OT

Sinkhorn

Unbalanced and Partial OT

Structured OT

Dynamic OT & Schrodinger Bridge

Multi-Marginal & Wasserstein Barycenter

Tree Wasserstein

Gromov Wasserstein

Sliced Wasserstein

Other formulations

Theoretical and Complexity

Scaling OT

Measures simplification

Cost and Kernel Approximation

Plan Structuring

Computation Partitioning

Solver Acceleration

Applications

Generative Models

Domain Adaptation

Other Applications

Software and Tools

Files

README.md

Latest commit

History

README.md

File metadata and controls

Scalable Optimal Transport

Contents

Background

Regularized OT

Sinkhorn

Unbalanced and Partial OT

Structured OT

Dynamic OT & Schrodinger Bridge

Multi-Marginal & Wasserstein Barycenter

Tree Wasserstein

Gromov Wasserstein

Sliced Wasserstein

Other formulations

Theoretical and Complexity

Scaling OT

Measures simplification

Cost and Kernel Approximation

Plan Structuring

Computation Partitioning

Solver Acceleration

Applications

Generative Models

Domain Adaptation

Other Applications

Software and Tools