Boosting Camera Motion Control for Video Diffusion Transformers

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cheong, Soon Yau, Ceylan, Duygu, Mustafa, Armin, Gilbert, Andrew, Huang, Chun-Hao Paul
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913545390653440
author Cheong, Soon Yau
Ceylan, Duygu
Mustafa, Armin
Gilbert, Andrew
Huang, Chun-Hao Paul
author_facet Cheong, Soon Yau
Ceylan, Duygu
Mustafa, Armin
Gilbert, Andrew
Huang, Chun-Hao Paul
contents Recent advancements in diffusion models have significantly enhanced the quality of video generation. However, fine-grained control over camera pose remains a challenge. While U-Net-based models have shown promising results for camera control, transformer-based diffusion models (DiT)-the preferred architecture for large-scale video generation - suffer from severe degradation in camera motion accuracy. In this paper, we investigate the underlying causes of this issue and propose solutions tailored to DiT architectures. Our study reveals that camera control performance depends heavily on the choice of conditioning methods rather than camera pose representations that is commonly believed. To address the persistent motion degradation in DiT, we introduce Camera Motion Guidance (CMG), based on classifier-free guidance, which boosts camera control by over 400%. Additionally, we present a sparse camera control pipeline, significantly simplifying the process of specifying camera poses for long videos. Our method universally applies to both U-Net and DiT models, offering improved camera control for video generation tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2410_10802
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Boosting Camera Motion Control for Video Diffusion Transformers
Cheong, Soon Yau
Ceylan, Duygu
Mustafa, Armin
Gilbert, Andrew
Huang, Chun-Hao Paul
Computer Vision and Pattern Recognition
Artificial Intelligence
Recent advancements in diffusion models have significantly enhanced the quality of video generation. However, fine-grained control over camera pose remains a challenge. While U-Net-based models have shown promising results for camera control, transformer-based diffusion models (DiT)-the preferred architecture for large-scale video generation - suffer from severe degradation in camera motion accuracy. In this paper, we investigate the underlying causes of this issue and propose solutions tailored to DiT architectures. Our study reveals that camera control performance depends heavily on the choice of conditioning methods rather than camera pose representations that is commonly believed. To address the persistent motion degradation in DiT, we introduce Camera Motion Guidance (CMG), based on classifier-free guidance, which boosts camera control by over 400%. Additionally, we present a sparse camera control pipeline, significantly simplifying the process of specifying camera poses for long videos. Our method universally applies to both U-Net and DiT models, offering improved camera control for video generation tasks.
title Boosting Camera Motion Control for Video Diffusion Transformers
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2410.10802