I3D 2025 Papers Session 3 - Animation & Voxels
Auto-discovered
Visit resource ↗Description
This I3D 2025 Papers Session 3 recording is a YouTube video from the I3D Symposium channel. Hosted by Paul Torrens, it presents four research talks under the Animation and Voxels theme.
The papers include hierarchical neural skinning deformation with self-supervised training for character animation, foveated animation updates for efficient crowd simulation, transform-aware sparse voxel directed acyclic graphs, and Aokana for real-time rendering of large-scale open-world voxels with LOD and a visibility buffer. Each talk covers motivation, method, results, and audience Q&A.
The session suits advanced researchers and engineers working on character deformation, crowd animation cost, or sparse voxel representations. Prior graphics research literacy is expected.
It is a free full session upload; individual papers are presented live with slides rather than as separate edited shorts. Session context and voting reminders appear around the talks.
It is useful as a compact view of current I3D work that bridges neural animation, perceptual LOD for crowds, and scalable voxel data structures.