Pretraining Frame Preservation in Autoregressive Video Memory Compression
Paper
•
2512.23851
•
Published
•
24
[SIGGRAPH2025] RenderFormer: Transformer-based Neural Rendering of Triangle Meshes with Global Illumination