Research map / Video / Video generation
AutoMV: An Autonomous Agent Framework for Real Estate Marketing Video Generation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Introduces AutoMV, an MLLM-agent framework that orchestrates a tool library (GPT-4V planning, Grounding DINO, SwinIR, MotionCtrl, RIFE, post-processing) to turn real estate images and text into marketing videos.
Problem
Real estate listings rely on static images; producing marketing videos is costly and complex for individual homeowners and small firms, and text-prompted video diffusion models give poor control over camera motion and storyline.
Approach
An MLLM agent decomposes the video-production task, generates storyline directives and camera trajectories, and autonomously selects and calls existing vision, super-resolution, camera-controlled video diffusion, interpolation and post-processing tools from a tool library.
Key results
- No quantitative results are reported; the paper describes the system architecture and a demo website with example generated videos.
Limitations
- No quantitative evaluation or ablation of the agent or tools is reported.
- Depends on third-party tools and closed GPT-4V for planning.
- Future work limited to expanding the tool library.
Experimental setup
Builds on: MotionCtrl Grounding DINO SwinIR RIFE
Closest papers by meaning
- AgenticVBench: Can AI Agents Complete Real-World Video Production Tasks? (NeurIPS 2026 · similarity 0.80)
- Position: Video as the New Language for Real-World Decision Making (ICML 2024 · similarity 0.78)
- EvoGround: Self-Evolving Video Agents for Video Temporal Grounding (NeurIPS 2026 · similarity 0.78)
- SMART: Scalable Multi-agent Real-time Motion Generation via Next-token Prediction (NeurIPS 2024 · similarity 0.78)
- AgentOccam: A Simple Yet Strong Baseline for LLM-Based Web Agents (ICLR 2025 · similarity 0.78)
- AgentStudio: A Toolkit for Building General Virtual Agents (ICLR 2025 · similarity 0.78)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.