---
title: ShutterMuse Video Demo
emoji: 📷
colorFrom: blue
colorTo: purple
sdk: static
app_file: index.html
pinned: false
models:
- ShutterMuse/ShutterMuse
datasets:
- ShutterMuse/CaptureGuide-Bench
tags:
- arxiv:2606.25763
- photography
- multimodal
- vision-language
- video-demo
---

# ShutterMuse Video Demo

This Space hosts the video demo for [ShutterMuse: Capture-Time Photography Guidance with MLLMs](https://huggingface.co/papers/2606.25763).

- **Paper:** https://arxiv.org/abs/2606.25763
- **Paper Page:** https://huggingface.co/papers/2606.25763
- **Project Page:** https://lijayuTnT.github.io/ShutterMuse/
- **GitHub Repository:** https://github.com/lijayuTnT/ShutterMuse
- **ShutterMuse Model:** https://huggingface.co/ShutterMuse/ShutterMuse
- **CaptureGuide-Bench:** https://huggingface.co/datasets/ShutterMuse/CaptureGuide-Bench

ShutterMuse is a unified multimodal large language model for capture-time photography guidance. It supports photographer-side composition recommendation and subject-side pose recommendation.