PhotoFlow: Agentic 3D Virtual Photography Missions

• URL: http://arxiv.org/abs/2605.23771v1
• Date: Fri, 22 May 2026 15:40:52 GMT
• Title: PhotoFlow: Agentic 3D Virtual Photography Missions
• Abstract summary: PhotoFlow is a Director-Reviewer-Reflector agent for closed-loop camera search.<n>PhotoFlow achieves the strongest external quality-alignment composite and success rate among one-shot prediction, single-chain reflection, anchor-bank selection, and random search.<n>This is the first work to make language-conditioned virtual photography in arbitrary Blender scenes an executable agent task.
• Score: 24.105785278409584
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: 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.

Related papers

• An Evaluation Framework for Generating Multi-View Images of a Person in a Scene [8.826327162415287]
  A primary bottleneck in training foundation models to execute free-form, promptable camera angle changes is the lack of specialized training data.<n>We propose the Head Scene Rotation Difference (HSRD) metric to quantitatively evaluate camera movements around a person.<n>HSRD provides the pipeline necessary to evaluate 3D spatial parallax for a person in a scene, paving the way to reliably construct high-quality multi-view synthetic datasets.
  arXiv  Detail & Related papers  (2026-09-04T01:02:12Z)
• CinemaTraj: Composing Atomic Camera Trajectories for 3D Scenes with LLM Agents [14.879115975370842]
  CinemaTraj is a framework that reframes camera trajectory planning as a language-grounded spatial reasoning problem.<n>It produces prompt-faithful, collision-free trajectories with high cinematographic quality.
  arXiv  Detail & Related papers  (2026-07-29T13:41:52Z)
• SimWorlds: A Multi-Agent System for Dynamic 3D Scene Creation [50.175987640001644]
  We present SimWorlds: a multi-agent framework that produces dynamic, editable 4D scenes from text.<n>We also introduce 4DBuildBench, a benchmark for assessing both visual fidelity and physical consistency of the procedural dynamic 3D scenes.
  arXiv  Detail & Related papers  (2026-07-02T06:29:18Z)
• Before the Shutter: Aesthetic and Actionable Portrait Photography Planning in 3D Scenes [35.590517071520964]
  We introduce 3D aesthetic portrait planning, the task of generating human pose, camera, lighting, and exposure plans.<n>Our approach builds a Photographic Scene Graph that represents scene affordances, subject-scene relations, and portrait-relevant lighting structure.<n>Our results suggest a path from post-capture correction toward pre-capture computational portrait planning.
  arXiv  Detail & Related papers  (2026-05-28T17:55:09Z)
• PhotoAgent: A Robotic Photographer with Spatial and Aesthetic Understanding [10.434774696873793]
  PhotoAgent translates subjective aesthetic goals into solvable geometric constraints.<n>This initial pose is then iteratively refined through visual reflection within a photorealistic internal world model.<n> Evaluations confirm that PhotoAgent excels in spatial reasoning and superior final image quality.
  arXiv  Detail & Related papers  (2026-03-24T04:40:33Z)
• HouseTour: A Virtual Real Estate A(I)gent [80.36635722117329]
  We introduce HouseTour, a method for spatially-aware 3D camera trajectory and natural language summary generation.<n>Our approach generates smooth video trajectories via a diffusion process constrained by known camera poses.<n>We synthesize the final video using 3D Gaussian splatting to render novel views along the trajectory.
  arXiv  Detail & Related papers  (2025-10-20T19:47:35Z)
• Lazy Visual Localization via Motion Averaging [89.8709956317671]
  We show that it is possible to achieve high localization accuracy without reconstructing the scene from the database. Experiments show that our visual localization proposal, LazyLoc, achieves comparable performance against state-of-the-art structure-based methods.
  arXiv  Detail & Related papers  (2023-07-19T13:40:45Z)
• Automatic Camera Trajectory Control with Enhanced Immersion for Virtual Cinematography [23.070207691087827]
  Real-world cinematographic rules show that directors can create immersion by comprehensively synchronizing the camera with the actor. Inspired by this strategy, we propose a deep camera control framework that enables actor-camera synchronization in three aspects. Our proposed method yields immersive cinematic videos of high quality, both quantitatively and qualitatively.
  arXiv  Detail & Related papers  (2023-03-29T22:02:15Z)
• Factorized and Controllable Neural Re-Rendering of Outdoor Scene for Photo Extrapolation [50.00344639039158]
  We propose a factorized neural re-rendering model to produce novel views from cluttered outdoor Internet photo collections. We also propose a novel realism augmentation process to complement appearance details, which automatically propagates the texture details from a narrow captured photo to the extrapolated neural rendered image.
  arXiv  Detail & Related papers  (2022-07-14T13:28:08Z)
• PhotoScene: Photorealistic Material and Lighting Transfer for Indoor Scenes [84.66946637534089]
  PhotoScene is a framework that takes input image(s) of a scene and builds a photorealistic digital twin with high-quality materials and similar lighting. We model scene materials using procedural material graphs; such graphs represent photorealistic and resolution-independent materials. We evaluate our technique on objects and layout reconstructions from ScanNet, SUN RGB-D and stock photographs, and demonstrate that our method reconstructs high-quality, fully relightable 3D scenes.
  arXiv  Detail & Related papers  (2022-07-02T06:52:44Z)
• 3D Moments from Near-Duplicate Photos [67.15199743223332]
  3D Moments is a new computational photography effect. We produce a video that smoothly interpolates the scene motion from the first photo to the second. Our system produces photorealistic space-time videos with motion parallax and scene dynamics.
  arXiv  Detail & Related papers  (2022-05-12T17:56:18Z)
• Environment Predictive Coding for Embodied Agents [92.31905063609082]
  We introduce environment predictive coding, a self-supervised approach to learn environment-level representations for embodied agents. Our experiments on the photorealistic 3D environments of Gibson and Matterport3D show that our method outperforms the state-of-the-art on challenging tasks with only a limited budget of experience.
  arXiv  Detail & Related papers  (2021-02-03T23:43:16Z)
• Batteries, camera, action! Learning a semantic control space for expressive robot cinematography [15.895161373307378]
  We develop a data-driven framework that enables editing of complex camera positioning parameters in a semantic space. First, we generate a database of video clips with a diverse range of shots in a photo-realistic simulator. We use hundreds of participants in a crowd-sourcing framework to obtain scores for a set of semantic descriptors for each clip.
  arXiv  Detail & Related papers  (2020-11-19T21:56:53Z)
• A simulation environment for drone cinematography [7.324046599137339]
  Methods for environmental image capture, 3D reconstruction (photogrammetry) and the creation of foreground assets are presented. The tool supports both free-flight and parameterisable standard shot types along with programmable scenarios associated with foreground assets and event dynamics.
  arXiv  Detail & Related papers  (2020-10-03T09:57:56Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.