| CARVIEW |
Select Language
HTTP/2 301
server: GitHub.com
content-type: text/html
location: https://hehao13.github.io/Projects-CameraCtrl-II/
x-github-request-id: 0D08:2D8B9D:871A8C:97D427:69522D77
accept-ranges: bytes
age: 0
date: Mon, 29 Dec 2025 07:27:52 GMT
via: 1.1 varnish
x-served-by: cache-bom-vanm7210084-BOM
x-cache: MISS
x-cache-hits: 0
x-timer: S1766993272.061400,VS0,VE201
vary: Accept-Encoding
x-fastly-request-id: 9779a0d54ead67543bffe596ea22f09e3d8a5af1
content-length: 162
HTTP/2 200
server: GitHub.com
content-type: text/html; charset=utf-8
last-modified: Fri, 14 Mar 2025 02:22:14 GMT
access-control-allow-origin: *
strict-transport-security: max-age=31556952
etag: W/"67d392d6-4c54"
expires: Mon, 29 Dec 2025 07:37:52 GMT
cache-control: max-age=600
content-encoding: gzip
x-proxy-cache: MISS
x-github-request-id: BF82:2680BD:86E2E8:979B9B:69522D78
accept-ranges: bytes
age: 0
date: Mon, 29 Dec 2025 07:27:52 GMT
via: 1.1 varnish
x-served-by: cache-bom-vanm7210084-BOM
x-cache: MISS
x-cache-hits: 0
x-timer: S1766993272.291987,VS0,VE205
vary: Accept-Encoding
x-fastly-request-id: 0a886412b0df378962e4458b5c403c6dc353d050
content-length: 3529
CameraCtrl II
CameraCtrl II: Dynamic Scene Exploration via Camera-controlled
Research Paper
CameraCtrl II: Dynamic Scene Exploration via Camera-controlled
Video Diffusion Models
Hao He1,2
Ceyuan Yang2,✝
Shanchuan Lin2
Yinghao Xu3
Meng Wei4
Liangke Gui2
Qi Zhao2
Gordon Wetzstein3
Lu Jiang2
Hongsheng Li1
1The Chinese University of Hong Kong
2ByteDance Seed
3Stanford University
4ByteDance
✝Corresponding Author
Dynamic Scenes Exploration
To enable broader exploration of dynamic scenes, our model can generate new video clips of the same scene based on previously generated content and user-provided camera trajectories. This approach maintains dynamic capabilities, accurate camera control, and scene consistency throughout the extended exploration.
Indoor exploration
Hotel Lobby Exploration
Outdoor exploration
Walking in Park
Exploration in different scenarios
Our model enables precise camera control across diverse scenarios while preserving dynamic scene elements, e.g.
City walking
European Road
Mediterranean Market Street
19th Century Foggy London
Foggy London Railway Station
Street View before Dawn
Fantasy Hiking Adventure
Mountain trail with flowing stream
Cliffside path with waterfall rainbow
Minecraft-Style Gaming Environment
Voxel village with windmills sunset
Voxel village with people, sheep
Realistic Indoor Settings
Old cozy library
Abandoned hospital
3D reconstruction of generated videos
Our method can generate videos with strong 3D consistency, which enables high-quality 3D reconstruction using the camera-controlled videos.
Generated Videos
Reconstruction Point Clouds