SOURCE-LINKED INTELLIGENCE

ABCD: Alpha-Composited Block Coordinate Descent: Constant-VRAM Training for Large Radiance Fields

arXiv · AI, language, vision and robotics · article · Aug 27, 2026 · UTC

We present ABCD (Alpha-Composited Block Coordinate Descent), an out-of-core training framework for alpha-composited radiance fields, instantiated here for 3D Gaussian Splatting. Our method reformulates training as block coordinate descent over spatial partitions: only one block of parameters is active at a time, while all others are frozen. By exploiting the associativity of alpha blending, these inactive regions can be pre-rendered and collapsed into foreground and background RGBA images. As a result, for fixed partition size and image resolution, peak VRAM becomes O(1) with respect to total

Read original source ↗ Open in workspace

recordType

paper

region

Global