DualCamCtrl: Dual-Branch Diffusion Model for Geometry-Aware Camera-Controlled Video Generation

TL;DR

DualCamCtrl is a new diffusion-based model that improves camera-controlled video generation by integrating RGB and depth data through a dual-branch framework and a semantic-guided alignment mechanism, enhancing scene understanding and geometric consistency.

Contribution

It introduces a dual-branch framework with SIGMA for better RGB-depth fusion, advancing scene understanding and geometric accuracy in camera-controlled video synthesis.

Findings

Achieves over 40% reduction in camera motion errors.

Generates more consistent and geometry-aware videos.

Effectively disentangles appearance and geometry modeling.

Abstract

This paper presents DualCamCtrl, a novel end-to-end diffusion model for camera-controlled video generation. Recent works have advanced this field by representing camera poses as ray-based conditions, yet they often lack sufficient scene understanding and geometric awareness. DualCamCtrl specifically targets this limitation by introducing a dual-branch framework that mutually generates camera-consistent RGB and depth sequences. To harmonize these two modalities, we further propose the Semantic Guided Mutual Alignment (SIGMA) mechanism, which performs RGB-depth fusion in a semantics-guided and mutually reinforced manner. These designs collectively enable DualCamCtrl to better disentangle appearance and geometry modeling, generating videos that more faithfully adhere to the specified camera trajectories. Additionally, we analyze and reveal the distinct influence of depth and camera poses across denoising stages and further demonstrate that early and late stages play complementary roles in forming global structure and refining local details. Extensive experiments demonstrate that DualCamCtrl achieves more consistent camera-controlled video generation, with over 40\% reduction in camera motion errors compared with prior methods. Our project page: https://soyouthinkyoucantell.github.io/dualcamctrl-page/