OmniFlow: Any-to-Any Generation with Multi-Modal Rectified Flows

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Shufan, Kallidromitis, Konstantinos, Gokul, Akash, Liao, Zichun, Kato, Yusuke, Kozuka, Kazuki, Grover, Aditya
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915207181238272
author Li, Shufan
Kallidromitis, Konstantinos
Gokul, Akash
Liao, Zichun
Kato, Yusuke
Kozuka, Kazuki
Grover, Aditya
author_facet Li, Shufan
Kallidromitis, Konstantinos
Gokul, Akash
Liao, Zichun
Kato, Yusuke
Kozuka, Kazuki
Grover, Aditya
contents We introduce OmniFlow, a novel generative model designed for any-to-any generation tasks such as text-to-image, text-to-audio, and audio-to-image synthesis. OmniFlow advances the rectified flow (RF) framework used in text-to-image models to handle the joint distribution of multiple modalities. It outperforms previous any-to-any models on a wide range of tasks, such as text-to-image and text-to-audio synthesis. Our work offers three key contributions: First, we extend RF to a multi-modal setting and introduce a novel guidance mechanism, enabling users to flexibly control the alignment between different modalities in the generated outputs. Second, we propose a novel architecture that extends the text-to-image MMDiT architecture of Stable Diffusion 3 and enables audio and text generation. The extended modules can be efficiently pretrained individually and merged with the vanilla text-to-image MMDiT for fine-tuning. Lastly, we conduct a comprehensive study on the design choices of rectified flow transformers for large-scale audio and text generation, providing valuable insights into optimizing performance across diverse modalities. The Code will be available at https://github.com/jacklishufan/OmniFlows.
format Preprint
id arxiv_https___arxiv_org_abs_2412_01169
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle OmniFlow: Any-to-Any Generation with Multi-Modal Rectified Flows
Li, Shufan
Kallidromitis, Konstantinos
Gokul, Akash
Liao, Zichun
Kato, Yusuke
Kozuka, Kazuki
Grover, Aditya
Multimedia
Computer Vision and Pattern Recognition
Sound
Audio and Speech Processing
We introduce OmniFlow, a novel generative model designed for any-to-any generation tasks such as text-to-image, text-to-audio, and audio-to-image synthesis. OmniFlow advances the rectified flow (RF) framework used in text-to-image models to handle the joint distribution of multiple modalities. It outperforms previous any-to-any models on a wide range of tasks, such as text-to-image and text-to-audio synthesis. Our work offers three key contributions: First, we extend RF to a multi-modal setting and introduce a novel guidance mechanism, enabling users to flexibly control the alignment between different modalities in the generated outputs. Second, we propose a novel architecture that extends the text-to-image MMDiT architecture of Stable Diffusion 3 and enables audio and text generation. The extended modules can be efficiently pretrained individually and merged with the vanilla text-to-image MMDiT for fine-tuning. Lastly, we conduct a comprehensive study on the design choices of rectified flow transformers for large-scale audio and text generation, providing valuable insights into optimizing performance across diverse modalities. The Code will be available at https://github.com/jacklishufan/OmniFlows.
title OmniFlow: Any-to-Any Generation with Multi-Modal Rectified Flows
topic Multimedia
Computer Vision and Pattern Recognition
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2412.01169