Presentation

<
SmoothMotionVectors: Optimizing Your Content for Video Codecs in Free View Video Compression
SessionDynamic Scenes
DescriptionConventional video codecs exploit spatial and temporal self-similarity to compress structured signals. We bring this principle to dynamic scene reconstruction with 3D Gaussian Splatting by organizing deformation outputs through a tailored pipeline. The resulting motion vectors are smoother across space and time, allowing greater storage reduction without sacrificing visual quality.
Authors
Mingyang Song
Disney Research Studios
ETH Zürich
Yang Zhang
Disney Research Studios
Siyu Tang
ETH Zürich
Tunc Ozan Aydin
Disney Research Studios
Event Type
Technical Paper
TimeTuesday, 21 July 202610:55am - 11:05am PDT
LocationRoom 408 B
Digital Library PDF
Session Time & Location
Sunday, 19 July 20266:00pm - 8:45pm PDTHall K
Tuesday, 21 July 202610:45am - 12:15pm PDTRoom 408 B
Keywords
Rendering
Registration Categories
Full Conference Supporter
Full Conference
Next PresentationNext PresentationCAGS: Color-Adaptive Volumetric Video Streaming with Dynamic 3D Gaussian Splatting
Similar Presentations
Implicit Surface Compression -- with Good Old Discrete Cosine Transform and Motion Compensation
LongE2V: Long-Horizon Event-based Video Reconstruction, Prediction, and Frame Interpolation with Video Diffusion Models
MV-S2V: Multi-View Subject-Consistent Video Generation