Exploring Recurrent Long-term Temporal Fusion for Multi-view 3D Perception

arXiv:2303.05970v3 Announce Type: replace
Abstract: Long-term temporal fusion is a crucial but often overlooked technique in camera-based Bird's-Eye-View (BEV) 3D perception. Existing methods are mostly in a parallel manner. While parallel fusion can benefit from long-term information, it suffers from increasing computational and memory overheads as the fusion window size grows. Alternatively, BEVFormer adopts a recurrent fusion pipeline so that history information can be efficiently integrated, yet it fails to benefit from longer temporal frames. In this paper, we explore …

abstract arxiv benefit bird computational cs.cv fusion information long-term memory perception temporal type view