InverseMatrixVT3D: An Efficient Projection Matrix-Based Approach for 3D Occupancy Prediction

arXiv:2401.12422v2 Announce Type: replace
Abstract: This paper introduces InverseMatrixVT3D, an efficient method for transforming multi-view image features into 3D feature volumes for 3D semantic occupancy prediction. Existing methods for constructing 3D volumes often rely on depth estimation, device-specific operators, or transformer queries, which hinders the widespread adoption of 3D occupancy models. In contrast, our approach leverages two projection matrices to store the static mapping relationships and matrix multiplications to efficiently generate global Bird's Eye View (BEV) features and local 3D …

arxiv cs.cv cs.ro matrix prediction projection type