SuBloNet: Sparse Super Block Networks for Large Scale Volumetric Fusion

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
The Eurographics Association
Training and inference of convolutional neural networks (CNNs) on truncated signed distance fields (TSDFs) is a challenging task. Large parts of the scene are usually empty, which makes dense implementations inefficient in terms of memory consumption and compute throughput. However, due to the truncation distance, non-zero values are grouped around the surface creating small dense blocks inside the large empty space. We show that this structure can be exploited by storing the TSDF in a block sparse tensor and then decomposing it into rectilinear super blocks. A super block is a dense 3d cuboid of variable size and can be processed by conventional CNNs. We analyze the rectilinear decomposition and present a formulation for computing the bandwidth-optimal solution given a specific network architecture. However, this solution is NP-complete, therefore we also a present a heuristic approach for fast training and inference tasks. We verify the effectiveness of SuBloNet and report a speedup of 4x towards dense implementations and 1.7x towards state-of-the-art sparse implementations. Using the super block architecture, we show that recurrent volumetric fusion is now possible on large scale scenes. Such a systems is able to reconstruct high-quality surfaces from few noisy depth images.

, booktitle = {
Vision, Modeling, and Visualization
}, editor = {
Andres, Bjoern and Campen, Marcel and Sedlmair, Michael
}, title = {{
SuBloNet: Sparse Super Block Networks for Large Scale Volumetric Fusion
}}, author = {
Rückert, Darius
Stamminger, Marc
}, year = {
}, publisher = {
The Eurographics Association
}, ISBN = {
}, DOI = {
} }