Diffusion & Adversarial Schrödinger Bridges via Iterative Proportional Markovian Fitting
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866912940580405248 |
|---|---|
| author | Kholkin, Sergei Ksenofontov, Grigoriy Li, David Kornilov, Nikita Gushchin, Nikita Suvorikova, Alexandra Kroshnin, Alexey Burnaev, Evgeny Korotin, Alexander |
| author_facet | Kholkin, Sergei Ksenofontov, Grigoriy Li, David Kornilov, Nikita Gushchin, Nikita Suvorikova, Alexandra Kroshnin, Alexey Burnaev, Evgeny Korotin, Alexander |
| contents | The Iterative Markovian Fitting (IMF) procedure, which iteratively projects onto the space of Markov processes and the reciprocal class, successfully solves the Schrödinger Bridge (SB) problem. However, an efficient practical implementation requires a heuristic modification -- alternating between fitting forward and backward time diffusion at each iteration. This modification is crucial for stabilizing training and achieving reliable results in applications such as unpaired domain translation. Our work reveals a close connection between the modified version of IMF and the Iterative Proportional Fitting (IPF) procedure -- a foundational method for the SB problem, also known as Sinkhorn's algorithm. Specifically, we demonstrate that the heuristic modification of the IMF effectively integrates both IMF and IPF procedures. We refer to this combined approach as the Iterative Proportional Markovian Fitting (IPMF) procedure. Through theoretical and empirical analysis, we establish the convergence of the IPMF procedure under various settings, contributing to developing a unified framework for solving SB problems. Moreover, from a practical standpoint, the IPMF procedure enables a flexible trade-off between image similarity and generation quality, offering a new mechanism for tailoring models to specific tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_02601 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Diffusion & Adversarial Schrödinger Bridges via Iterative Proportional Markovian Fitting Kholkin, Sergei Ksenofontov, Grigoriy Li, David Kornilov, Nikita Gushchin, Nikita Suvorikova, Alexandra Kroshnin, Alexey Burnaev, Evgeny Korotin, Alexander Machine Learning The Iterative Markovian Fitting (IMF) procedure, which iteratively projects onto the space of Markov processes and the reciprocal class, successfully solves the Schrödinger Bridge (SB) problem. However, an efficient practical implementation requires a heuristic modification -- alternating between fitting forward and backward time diffusion at each iteration. This modification is crucial for stabilizing training and achieving reliable results in applications such as unpaired domain translation. Our work reveals a close connection between the modified version of IMF and the Iterative Proportional Fitting (IPF) procedure -- a foundational method for the SB problem, also known as Sinkhorn's algorithm. Specifically, we demonstrate that the heuristic modification of the IMF effectively integrates both IMF and IPF procedures. We refer to this combined approach as the Iterative Proportional Markovian Fitting (IPMF) procedure. Through theoretical and empirical analysis, we establish the convergence of the IPMF procedure under various settings, contributing to developing a unified framework for solving SB problems. Moreover, from a practical standpoint, the IPMF procedure enables a flexible trade-off between image similarity and generation quality, offering a new mechanism for tailoring models to specific tasks. |
| title | Diffusion & Adversarial Schrödinger Bridges via Iterative Proportional Markovian Fitting |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2410.02601 |