CamPilot improves text to movie creation with cinematic camera control
CamPilot: A Multi-Agent Cinematic Assistant for Camera-Controlled Movie Generation
Computer Vision and Pattern Recognition
Summary
Making movies with computers is still hard because camera movements and shots often look less natural than those made by human filmmakers. The authors created CamPilot, a system that learns from thousands of professional movies how to plan camera angles and motions in ways that look artistic and natural. It uses multiple AI agents working together to control the cameras and keep scenes connected smoothly. Their method results in videos that look more like real movies than previous automatic systems.
What this means in practice
- •For video game developers: Automate realistic and artistic in-game camera movements to enhance storytelling and player immersion.
- •For advertising agencies: Generate promotional videos with professional camera work style automatically from text scripts to speed up production.
Authors
Yang Wu, Stefano Petrangeli, Ishita Dasgupta, Yu Shen
Abstract
The integration of large language models (LLMs) into video generation has enabled rapid text-to-video creation and improved visual quality. However, it still falls short of professional filmmaking, where cinematographic language is less refined than human-crafted camera work and multi-shot continuity remains challenging. To address these limitations, we introduce CamPilot, a multi-agent framework that integrates cinematographic planning and camera-work control to produce more coherent, logically structured, and human-aesthetic movies. CamPilot adopts a GRPO-based learning paradigm to learn camera work planning from 14K real-world professional movies, internalizing motion patterns and composition principles that support reasoning over shooting techniques (e.g., camera angle, motion, and focal behavior) and cross-shot relationships for controllable camera-viewpoint generation. Multiple agents further collaborate and evolve to improve overall output quality. To support this work and further studies in this domain, we establish CamEval, a benchmark for evaluating camera work quality and cinematic engagement. Empirical results show that CamPilot outperforms state-of-the-art text-to-movie generation methods on cinematographic control and quality, highlighting the impact of professional camera design on movie generation.