Rectified flow learns straight-path neural ODEs for distribution transport, yielding efficient generative models and domain transfers that work well even with a single simulation step.
Likelihood training of schr \” odinger bridge using forward-backward sdes theory
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
SB-RF integrates SB time marginals with RF velocity objectives to train a conditional velocity field that performs single Euler-step enhancement from noisy speech, showing competitive results on VoiceBank-DEMAND and superior results on low-SNR tests.
A forward-backward HJB duality computes the optimal stochastic transport control from easy forward relaxation trajectories alone, expressed as path-space free energy without backward simulation.
Stochastic image enhancement methods are shown to be variants of a shared SDE differing in drift, diffusion, terminal distributions and boundary conditions, with controlled experiments revealing no single dominant family and a new modular library released.
MedShift applies flow matching and Schrödinger bridges for class-conditional unpaired translation between synthetic and real skull X-rays, benchmarked on the new X-DigiSkull dataset.
citing papers explorer
-
Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow
Rectified flow learns straight-path neural ODEs for distribution transport, yielding efficient generative models and domain transfers that work well even with a single simulation step.
-
SB-RF: Schr\"odinger Bridge Rectified Flow for One-Step Robust Speech Enhancement
SB-RF integrates SB time marginals with RF velocity objectives to train a conditional velocity field that performs single Euler-step enhancement from noisy speech, showing competitive results on VoiceBank-DEMAND and superior results on low-SNR tests.
-
Generative optimal transport via forward-backward HJB matching
A forward-backward HJB duality computes the optimal stochastic transport control from easy forward relaxation trajectories alone, expressed as path-space free energy without backward simulation.
-
Unifying Deep Stochastic Processes for Image Enhancement
Stochastic image enhancement methods are shown to be variants of a shared SDE differing in drift, diffusion, terminal distributions and boundary conditions, with controlled experiments revealing no single dominant family and a new modular library released.
-
MedShift: Implicit Conditional Transport for X-Ray Domain Adaptation
MedShift applies flow matching and Schrödinger bridges for class-conditional unpaired translation between synthetic and real skull X-rays, benchmarked on the new X-DigiSkull dataset.
- The Ensemble Schr{\"o}dinger Bridge filter for Nonlinear Data Assimilation