Exploring Recurrent Long-term Temporal Fusion for Multi-view 3D Perception
This work addresses computational and memory overheads in multi-view 3D perception for autonomous driving, offering an incremental improvement over existing methods.
The paper tackles the problem of inefficient long-term temporal fusion in camera-based Bird's-Eye-View 3D perception by proposing VideoBEV, a recurrent fusion strategy that achieves strong performance, such as 55.4% mAP for object detection and 48.6% vehicle mIoU for segmentation on the nuScenes benchmark.
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 an embarrassingly simple long-term recurrent fusion strategy built upon the LSS-based methods and find it already able to enjoy the merits from both sides, i.e., rich long-term information and efficient fusion pipeline. A temporal embedding module is further proposed to improve the model's robustness against occasionally missed frames in practical scenarios. We name this simple but effective fusing pipeline VideoBEV. Experimental results on the nuScenes benchmark show that VideoBEV obtains strong performance on various camera-based 3D perception tasks, including object detection (55.4\% mAP and 62.9\% NDS), segmentation (48.6\% vehicle mIoU), tracking (54.8\% AMOTA), and motion prediction (0.80m minADE and 0.463 EPA).