DITTO-2: Distilled Diffusion Inference-Time T-Optimization for Music Generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Novack, Zachary, McAuley, Julian, Berg-Kirkpatrick, Taylor, Bryan, Nicholas
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910464525467648
author Novack, Zachary
McAuley, Julian
Berg-Kirkpatrick, Taylor
Bryan, Nicholas
author_facet Novack, Zachary
McAuley, Julian
Berg-Kirkpatrick, Taylor
Bryan, Nicholas
contents Controllable music generation methods are critical for human-centered AI-based music creation, but are currently limited by speed, quality, and control design trade-offs. Diffusion Inference-Time T-optimization (DITTO), in particular, offers state-of-the-art results, but is over 10x slower than real-time, limiting practical use. We propose Distilled Diffusion Inference-Time T -Optimization (or DITTO-2), a new method to speed up inference-time optimization-based control and unlock faster-than-real-time generation for a wide-variety of applications such as music inpainting, outpainting, intensity, melody, and musical structure control. Our method works by (1) distilling a pre-trained diffusion model for fast sampling via an efficient, modified consistency or consistency trajectory distillation process (2) performing inference-time optimization using our distilled model with one-step sampling as an efficient surrogate optimization task and (3) running a final multi-step sampling generation (decoding) using our estimated noise latents for best-quality, fast, controllable generation. Through thorough evaluation, we find our method not only speeds up generation over 10-20x, but simultaneously improves control adherence and generation quality all at once. Furthermore, we apply our approach to a new application of maximizing text adherence (CLAP score) and show we can convert an unconditional diffusion model without text inputs into a model that yields state-of-the-art text control. Sound examples can be found at https://ditto-music.github.io/ditto2/.
format Preprint
id arxiv_https___arxiv_org_abs_2405_20289
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle DITTO-2: Distilled Diffusion Inference-Time T-Optimization for Music Generation
Novack, Zachary
McAuley, Julian
Berg-Kirkpatrick, Taylor
Bryan, Nicholas
Sound
Artificial Intelligence
Machine Learning
Controllable music generation methods are critical for human-centered AI-based music creation, but are currently limited by speed, quality, and control design trade-offs. Diffusion Inference-Time T-optimization (DITTO), in particular, offers state-of-the-art results, but is over 10x slower than real-time, limiting practical use. We propose Distilled Diffusion Inference-Time T -Optimization (or DITTO-2), a new method to speed up inference-time optimization-based control and unlock faster-than-real-time generation for a wide-variety of applications such as music inpainting, outpainting, intensity, melody, and musical structure control. Our method works by (1) distilling a pre-trained diffusion model for fast sampling via an efficient, modified consistency or consistency trajectory distillation process (2) performing inference-time optimization using our distilled model with one-step sampling as an efficient surrogate optimization task and (3) running a final multi-step sampling generation (decoding) using our estimated noise latents for best-quality, fast, controllable generation. Through thorough evaluation, we find our method not only speeds up generation over 10-20x, but simultaneously improves control adherence and generation quality all at once. Furthermore, we apply our approach to a new application of maximizing text adherence (CLAP score) and show we can convert an unconditional diffusion model without text inputs into a model that yields state-of-the-art text control. Sound examples can be found at https://ditto-music.github.io/ditto2/.
title DITTO-2: Distilled Diffusion Inference-Time T-Optimization for Music Generation
topic Sound
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2405.20289