arXiv cs.AI· Rohit Verma, Anand V Bodas·· 9 天前AI 评分27
BEVPIPE：用可移植 GPU 计算部署 BEV 感知管线，解决算子不匹配问题
The Operator Mismatch Problem: Deploying BEV Perception with Portable GPU Compute
AI 导读
BEVPIPE 是一个用可移植 GPU 计算 API 部署多模态 BEV 感知管线并接入生产推理运行时的框架，通过将模型拆分为运行时管理的稠密子图与体素化器、稀疏编码器、BEV 投影器三个外部算子扩展，并以共享 GPU 内存空间连接，实现相比传统部署 19.5 倍的端到端加速，同时保留 98.5% 的参考 mAP，并可跨不同 GPU 后端移植。
正文 · 原文
Abstract:Modern autonomous driving systems rely on bird's-eye-view (BEV) perception models that fuse camera and LiDAR inputs to detect objects in 3D space. These models are accurate, but they cannot be deployed through standard inference runtimes. The reason is an operator mismatch between dense convolutions (which runtimes handle well), sparse 3D convolutions (which runtimes cannot represent), and geometric scatter operations (which runtimes have no vocabulary for). Today, every sparse convolution library is CUDA-only and PyTorch-coupled, locking BEV deployment to a single vendor's hardware and a single execution framework.
We present BEVPIPE, a framework for deploying multimodal BEV perception pipelines using portable GPU compute APIs and integrating them with production inference runtimes. BEVPIPE partitions the model into runtime-managed dense subgraphs and three external operator extensions (voxelizer, sparse encoder, BEV projector), connected through a shared GPU memory space. BEVPIPE achieves a 19.5x end-to-end speedup over conventional deployments while retaining 98.5% of reference mAP. We also showcase that BEVPIPE is portable across different GPU backends.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.11504 [cs.AI] |
| (or arXiv:2610.11504v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.11504 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Rohit Verma [view email]
[v1]
Thu, 8 Oct 2026 08:42:45 UTC (57 KB)
来源：arXiv cs.AI · arxiv.org