CT-1: Vision-Language-Camera Models Transfer Spatial Reasoning Knowledge to Camera-Controllable Video Generation

TL;DR

This paper introduces CT-1, a novel vision-language-camera model that transfers spatial reasoning to generate camera-controllable videos with high accuracy, leveraging a large dataset and a frequency domain regularization technique.

Contribution

The paper presents a new model, CT-1, which effectively transfers spatial reasoning to video generation, and introduces CT-200K, a large-scale dataset for training such models.

Findings

Improved camera control accuracy by 25.7% over previous methods.

Successfully generates high-quality, spatially aware camera-controllable videos.

Employs Wavelet-based Regularization Loss to learn complex camera trajectories.

Abstract

Camera-controllable video generation aims to synthesize videos with flexible and physically plausible camera movements. However, existing methods either provide imprecise camera control from text prompts or rely on labor-intensive manual camera trajectory parameters, limiting their use in automated scenarios. To address these issues, we propose a novel Vision-Language-Camera model, termed CT-1 (Camera Transformer 1), a specialized model designed to transfer spatial reasoning knowledge to video generation by accurately estimating camera trajectories. Built upon vision-language modules and a Diffusion Transformer model, CT-1 employs a Wavelet-based Regularization Loss in the frequency domain to effectively learn complex camera trajectory distributions. These trajectories are integrated into a video diffusion model to enable spatially aware camera control that aligns with user intentions. To facilitate the training of CT-1, we design a dedicated data curation pipeline and construct CT-200K, a large-scale dataset containing over 47M frames. Experimental results demonstrate that our framework successfully bridges the gap between spatial reasoning and video synthesis, yielding faithful and high-quality camera-controllable videos and improving camera control accuracy by 25.7% over prior methods.