Variational Flow Models: Flowing in Your Style
- URL: http://arxiv.org/abs/2402.02977v3
- Date: Fri, 29 Mar 2024 12:28:46 GMT
- Title: Variational Flow Models: Flowing in Your Style
- Authors: Kien Do, Duc Kieu, Toan Nguyen, Dang Nguyen, Hung Le, Dung Nguyen, Thin Nguyen,
- Abstract summary: We transform the posterior flow of a "linear" process into a straight constant-speed (SC) flow, reminiscent of Rectified Flow.
This transformation facilitates fast sampling along the original posterior flow without training a new model of the SC flow.
We can easily integrate high-order numerical solvers into the transformed SC flow, further enhancing sampling accuracy and efficiency.
- Score: 32.913511518425864
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: We introduce "posterior flows" - generalizations of "probability flows" to a broader class of stochastic processes not necessarily diffusion processes - and propose a systematic training-free method to transform the posterior flow of a "linear" stochastic process characterized by the equation Xt = at * X0 + st * X1 into a straight constant-speed (SC) flow, reminiscent of Rectified Flow. This transformation facilitates fast sampling along the original posterior flow without training a new model of the SC flow. The flexibility of our approach allows us to extend our transformation to inter-convert two posterior flows from distinct "linear" stochastic processes. Moreover, we can easily integrate high-order numerical solvers into the transformed SC flow, further enhancing sampling accuracy and efficiency. Rigorous theoretical analysis and extensive experimental results substantiate the advantages of our framework.
Related papers
- Optimal Flow Matching: Learning Straight Trajectories in Just One Step [89.37027530300617]
We develop and theoretically justify the novel Optimal Flow Matching approach.
It allows recovering the straight OT displacement for the quadratic transport in just one FM step.
The main idea of our approach is the employment of vector field for FM which are parameterized by convex functions.
arXiv Detail & Related papers (2024-03-19T19:44:54Z) - Guided Flows for Generative Modeling and Decision Making [55.42634941614435]
We show that Guided Flows significantly improves the sample quality in conditional image generation and zero-shot text synthesis-to-speech.
Notably, we are first to apply flow models for plan generation in the offline reinforcement learning setting ax speedup in compared to diffusion models.
arXiv Detail & Related papers (2023-11-22T15:07:59Z) - Restoration-Degradation Beyond Linear Diffusions: A Non-Asymptotic
Analysis For DDIM-Type Samplers [90.45898746733397]
We develop a framework for non-asymptotic analysis of deterministic samplers used for diffusion generative modeling.
We show that one step along the probability flow ODE can be expressed as two steps: 1) a restoration step that runs ascent on the conditional log-likelihood at some infinitesimally previous time, and 2) a degradation step that runs the forward process using noise pointing back towards the current gradient.
arXiv Detail & Related papers (2023-03-06T18:59:19Z) - Flow Straight and Fast: Learning to Generate and Transfer Data with
Rectified Flow [32.459587479351846]
We present rectified flow, a surprisingly simple approach to learning (neural) ordinary differential equation (ODE) models.
We show that rectified flow performs superbly on image generation, image-to-image translation, and domain adaptation.
arXiv Detail & Related papers (2022-09-07T08:59:55Z) - GMFlow: Learning Optical Flow via Global Matching [124.57850500778277]
We propose a GMFlow framework for learning optical flow estimation.
It consists of three main components: a customized Transformer for feature enhancement, a correlation and softmax layer for global feature matching, and a self-attention layer for flow propagation.
Our new framework outperforms 32-iteration RAFT's performance on the challenging Sintel benchmark.
arXiv Detail & Related papers (2021-11-26T18:59:56Z) - Sinusoidal Flow: A Fast Invertible Autoregressive Flow [0.0]
We propose a new type of normalising flows that inherits the expressive power and triangular Jacobian from fully autoregressive flows.
Experiments show that our Sinusoidal Flow is not only able to model complex distributions, but can also be reliably inverted to generate realistic-looking samples.
arXiv Detail & Related papers (2021-10-26T01:20:29Z) - Towards extraction of orthogonal and parsimonious non-linear modes from
turbulent flows [0.0]
We propose a deep probabilistic-neural-network architecture for learning a minimal and near-orthogonal set of non-linear modes.
Our approach is based on $beta$-variational autoencoders ($beta$-VAEs) and convolutional neural networks (CNNs)
arXiv Detail & Related papers (2021-09-03T13:38:51Z) - SurVAE Flows: Surjections to Bridge the Gap between VAEs and Flows [78.77808270452974]
SurVAE Flows is a modular framework for composable transformations that encompasses VAEs and normalizing flows.
We show that several recently proposed methods, including dequantization and augmented normalizing flows, can be expressed as SurVAE Flows.
arXiv Detail & Related papers (2020-07-06T13:13:22Z) - Towards Recurrent Autoregressive Flow Models [39.25035894474609]
We present Recurrent Autoregressive Flows as a method toward general process modeling with normalizing flows.
The proposed method defines a conditional distribution for each variable in a sequential process by conditioning the parameters of a normalizing flow with recurrent neural connections.
We demonstrate the effectiveness of this class of models through a series of experiments in which models are trained on three complex processes.
arXiv Detail & Related papers (2020-06-17T18:38:36Z) - Gaussianization Flows [113.79542218282282]
We propose a new type of normalizing flow model that enables both efficient iteration of likelihoods and efficient inversion for sample generation.
Because of this guaranteed expressivity, they can capture multimodal target distributions without compromising the efficiency of sample generation.
arXiv Detail & Related papers (2020-03-04T08:15:06Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.