Schrödinger Bridge
A stochastic transport model describing the most likely stochastic evolution path given a source distribution and a target distribution. It can model transport between any two distributions and is widely used in conditional generation tasks such as image translation and domain adaptation.
Timeline (last 90 days)
The Schrödinger bridge (SB) model is a stochastic transport model that can build a bridge between a general source distribution and a target distribution.
Before BridgeTwist appeared, SB models lacked a training-free conditional sampling scheme.
BridgeTwist provides a training-free conditional sampler for pretrained SB models.
Diffusion models, flow models and SB models are the three mainstream frameworks of current generative models.
All Facts (4)
The Schrödinger bridge (SB) model is a stochastic transport model that can build a bridge between a general source distribution and a target distribution.
50%UnverifiedBefore BridgeTwist appeared, SB models lacked a training-free conditional sampling scheme.
50%UnverifiedBridgeTwist provides a training-free conditional sampler for pretrained SB models.
50%UnverifiedDiffusion models, flow models and SB models are the three mainstream frameworks of current generative models.
50%