Recent advancements in 3D generation have leveraged synthetic datasets with ground truth 3D assets and predefined camera trajectories. However, the potential of adopting real-world datasets, which can produce significantly more realistic 3D scenes, remains largely unexplored. In this work, we delve into the key challenge of the complex and scene-specific camera trajectories found in real-world captures. We introduce Director3D, a robust open-world text-to-3D generation framework, designed to generate both real-world 3D scenes and adaptive camera trajectories. To achieve this, (1) we first utilize a Trajectory Diffusion Transformer, acting as the Cinematographer, to model the distribution of camera trajectories based on textual descriptions. Next, a Gaussian-driven Multi-view Latent Diffusion Model serves as the Decorator, modeling the image sequence distribution given the camera trajectories and texts. This model, fine-tuned from a 2D diffusion model, directly generates pixel-aligned 3D Gaussians as an immediate 3D scene representation for consistent denoising. Lastly, the 3D Gaussians are further refined by a novel SDS++ loss as the Detailer, which incorporates the prior of the 2D diffusion model. Extensive experiments demonstrate that Director3D outperforms existing methods, offering superior performance in real-world 3D generation.
Our Director3D can handle different types of scenes with a joint framework. See more in our Gallery.
A white swan on a tranquil lake
A collection of fresh vegetables...
a chair that looks like fruit
A steep gorge carved by ancient rivers...
A snowy woodland path lined...
@article{li2024director3d,
author = {Xinyang Li and Zhangyu Lai and Linning Xu and Yansong Qu and Liujuan Cao and Shengchuan Zhang and Bo Dai and Rongrong Ji},
title = {Director3D: Real-world Camera Trajectory and 3D Scene Generation from Text},
journal = {arXiv:2406.17601},
year = {2024},
}