WorldCast: World-Consistent Character Video Generation with Large-Angle Camera Control
Summary
WorldCast 通过 3D 点云引导与几何对齐记忆注意力(GAMA),实现大角度相机运动下角色视频生成的外观、几何与身份一致性,并提出大规模基准数据集 CLACT。
View Cached Full Text
Cached at: 10/28/26, 01:38 AM
Paper page - WorldCast: World-Consistent Character Video Generation with Large-Angle Camera Control
Source: https://huggingface.co/papers/2610.33135
Abstract
CLACT and WorldCast enable consistent character video generation under large-angle camera motions by combining 3D point-cloud guidance with geometry-aligned memory attention.
Precise camera control is essential for cinematic character video generation, yet existing methods struggle to follow large-angle trajectories with combined rotation, translation, and radius variation from arbitrary initial azimuths. Such viewpoint changes also reveal character regions unseen in the initial frame, causing geometric distortions, appearance degradation, and identity drift. To systematically study these coupled challenges, we introduceCLACT, a large-scale Character-centric dataset and benchmark covering diverse Large-Angle Camera Trajectories, with accurate camera geometry and multi-view character observations for training and evaluation. We further proposeWorldCast, a framework for world-consistent character video generation under diverse large-angle camera motions. Specifically, we back-project the initial frame into a3D point cloudusing estimated depth, align the point-cloud scale with the targetcamera trajectory, and render it along the trajectory to obtain explicit geometric guidance and coverage masks. To complement the incomplete character information, we introduceGeometry-Aligned Memory Attention(GAMA). Usingrelative camera posesestimated jointly from the initial frame andmulti-view character anchors, we express anchor and target-view geometry in a common coordinate frame.GAMAuses this geometric alignment to guide the retrieval and injection of view-relevant character features during generation. Evaluations across diverse large-angle camera trajectories demonstrate accurate trajectory following while preserving character geometry, appearance, and identity.
View arXiv pageView PDFProject pageGitHub36Add to collection
Get this paper in your agent:
hf papers read 2610\.33135
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2610.33135 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2610.33135 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2610.33135 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
WorldCrafter: Consistent Video World Model with Implicit 3D-aware Memory
WorldCrafter is a video world model that learns a camera-queryable implicit 3D-aware memory for consistent and camera-controllable streaming scene exploration from a single image or text prompt.
CG-World: A Large-Scale World-State Dataset and Protocol for World Models
CG-World is a large-scale world-state dataset and protocol derived from industrial computer graphics pipelines, explicitly recording multimodal world states, interventions, and counterfactual branches to support world model research. It demonstrates improvements in geometry-conditioned video generation, action prediction, and closed-loop transfer of vision-language-action policies.
OmniCam: Omni-Camera Trajectory Generation via Geometry-Grounded Pose Token Learning
OmniCam 提出一种自回归模型,可从全景图与文本描述生成相机位姿序列,通过几何锚定的姿态 token 学习在 OmniCaT 数据集上将轨迹误差降低 28–47%、碰撞率降低 65.8%,并应用于相机可控视频生成与机器人主动感知。
WorldRecon: Generative Multi-View Mesh Reconstruction for Large-Scale 3D World
WorldRecon 提出一种生成式多视角网格重建框架,可从无位姿 RGB 图像重建大规模 3D 场景的完整网格。它将 TRELLIS.2 生成模型与重建式条件架构结合,通过邻域感知潜在修补和因果世界生成保证场景级一致性,在合成与真实室内场景上分别将 Chamfer 距离降低 14.9% 和 43.1%,达到最先进性能。
CausalAlign: Causal Alignment of Video Generation Models for Consistent Ego-Motion
CausalAlign 论文提出一种因果对齐框架,通过将扩散视频生成模型的内部表征与 3D 基础模型在 delta-token 空间中对齐,改善长时程相机自运动的一致性,在几何一致性上提升超过 20%,并提升了具身操作任务中世界动作模型的成功率。