# PhotoFlow: Agentic 3D Virtual Photography Missions

arXiv: https://arxiv.org/abs/2605.23771 · HTML: https://mlsanity.com/paper/2605.23771/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Multimodal and embodied agents, one of 16 directions in LLM agents and tool use.
- Builds on 23 of its 33 references that are in the map.
- First posted on arXiv 22 May 2026.

## Placed in

- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

## Builds on (most cited first, 23 of 23)

- [Matterport3D: Learning from RGB-D Data in Indoor Environments](https://mlsanity.com/paper/1709.06158/) (2017)
- [Habitat: A Platform for Embodied AI Research](https://mlsanity.com/paper/1904.01201/) (2019)
- [Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environments](https://mlsanity.com/paper/1711.07280/) (2017)
- [BLINK: Multimodal Large Language Models Can See but Not Perceive](https://mlsanity.com/paper/2404.12390/) (2024)
- [On Evaluation of Embodied Navigation Agents](https://mlsanity.com/paper/1807.06757/) (2018)
- [Gibson Env: Real-World Perception for Embodied Agents](https://mlsanity.com/paper/1808.10654/) (2018)
- [NIMA: Neural Image Assessment](https://mlsanity.com/paper/1709.05424/) (2017)
- [Visual Spatial Reasoning](https://mlsanity.com/paper/2205.00363/) (2022)
- [NavGPT: Explicit Reasoning in Vision-and-Language Navigation with Large Language Models](https://mlsanity.com/paper/2305.16986/) (2023)
- [What's "up" with vision-language models? Investigating their struggle with spatial reasoning](https://mlsanity.com/paper/2310.19785/) (2023)
- [Stay on the Path: Instruction Fidelity in Vision-and-Language Navigation](https://mlsanity.com/paper/1905.12255/) (2019)
- [Personalized Image Aesthetics Assessment with Rich Attributes](https://mlsanity.com/paper/2203.16754/) (2022)
- [FilmAgent: A Multi-Agent Framework for End-to-End Film Automation in Virtual 3D Spaces](https://mlsanity.com/paper/2501.12909/) (2025)
- [UniPercept: Towards Unified Perceptual-Level Image Understanding across Aesthetics, Quality, Structure, and Texture](https://mlsanity.com/paper/2512.21675/) (2025)
- [Autonomous Aerial Cinematography In Unstructured Environments With Learned Artistic Decision-Making](https://mlsanity.com/paper/1910.06988/) (2019)
- [ChatCam: Empowering Camera Control through Conversational AI](https://mlsanity.com/paper/2409.17331/) (2024)
- [AutoPhoto: Aesthetic Photo Capture using Reinforcement Learning](https://mlsanity.com/paper/2109.09923/) (2021)
- [Creatism: A deep-learning photographer capable of creating professional work](https://mlsanity.com/paper/1707.03491/) (2017)
- [CineMPC: A Fully Autonomous Drone Cinematography System Incorporating Zoom, Focus, Pose, and Scene Composition](https://mlsanity.com/paper/2401.05272/) (2024)
- [LeRoP: A Learning-Based Modular Robot Photography Framework](https://mlsanity.com/paper/1911.12470/) (2019)
- [Agentic Aerial Cinematography: From Dialogue Cues to Cinematic Trajectories](https://mlsanity.com/paper/2509.16176/) (2025)
- [Mind-of-Director: Multi-modal Agent-Driven Film Previsualization via Collaborative Decision-Making](https://mlsanity.com/paper/2603.14790/) (2026)
- [Aesthetic Camera Viewpoint Suggestion with 3D Aesthetic Field](https://mlsanity.com/paper/2602.20363/) (2026)

## Abstract (arXiv)

Virtual photography asks an agent to enter a prepared 3D scene with no preselected camera pose or reference image, infer a suitable shot from scene information and a language intent, choose executable camera parameters, and render the final photograph. Recent progress in vision-language models makes this kind of spatial agent increasingly plausible, but the task stresses two capabilities that remain hard to evaluate together: complex 3D spatial understanding and abstract aesthetic judgment. We introduce PhotoFlow, a Director-Reviewer-Reflector agent for closed-loop camera search. The Director builds a soft photographic blueprint and proposes diverse candidate cameras; the Reviewer combines rule checks, visual critique, and pairwise incumbent selection; and the Reflector converts failures into region memory, dead-zone suppression, and high-explore relocation. We also introduce VPhotoBench, a benchmark of 47 open-license Blender scenes and 141 language-conditioned photography missions spanning subject placement, relational composition, and atmosphere/style. On held-out experiments, PhotoFlow achieves the strongest external quality-alignment composite and success rate among one-shot prediction, single-chain reflection, anchor-bank selection, and random search under a six-round rendering budget. To our knowledge, this is the first work to make language-conditioned virtual photography in arbitrary Blender scenes an executable agent task, and our results show that an LLM-centered spatial agent can already produce strong photographs in a setting designed to challenge both 3D reasoning and aesthetic choice.