一句话：提出算法-硬件协同设计框架 AGS，利用视频 CODEC 辅助的帧共视检测加速 3DGS-SLAM，最高提速 17.12 倍。
3D 高斯泼溅（3DGS）与 SLAM 结合可取得出色的重建保真度，但每帧需要多次训练迭代、且高斯数量庞大，导致现有 3DGS-SLAM 系统吞吐不足。本文提出 AGS——一个算法与硬件协同设计框架，其核心直觉是 SLAM 以流式方式处理视频帧，相邻帧高度相似，这种相似性可用于加速。软件层面，作者针对机器人运动提出由粗到细的姿态跟踪方法，并通过跨帧共享高斯贡献信息避免冗余计算。硬件层面，作者设计帧共视检测引擎，从视频 CODEC 中提取中间数据来判定帧间共视关系，同时实现姿态跟踪引擎与建图引擎，并配合负载调度器高效部署上述算法。实验表明，AGS 相对移动端 GPU、高端 GPU 以及当前最先进的 3DGS 加速器 GSCore 分别取得最高 17.12×、6.71× 与 5.41× 的加速。
Simultaneous Localization and Mapping (SLAM) is a critical task that enables autonomous vehicles to construct maps and localize themselves in unknown environments. Recent breakthroughs combine SLAM with 3D Gaussian Splatting (3DGS) to achieve exceptional reconstruction fidelity. However, existing 3DGS-SLAM systems provide insufficient throughput due to the need for multiple training iterations per frame and the vast number of Gaussians. In this paper, we propose AGS, an algorithm-hardware co-design framework to boost the efficiency of 3DGS-SLAM based on the intuition that SLAM systems process frames in a streaming manner, where adjacent frames exhibit high similarity that can be utilized for acceleration. On the software level: 1) We propose a coarse-then-fine-grained pose tracking method with respect to the robot's movement. 2) We avoid redundant computations of Gaussians by sharing their contribution information across frames. On the hardware level, we propose a frame covisibility detection engine to extract intermediate data from the video CODEC. We also implement a pose tracking engine and a mapping engine with workload schedulers to efficiently deploy the AGS algorithm. Our evaluation shows that AGS achieves up to 17.12×, 6.71×, and 5.41× speedups against the mobile and high-end GPUs, and a state-of-the-art 3DGS accelerator, GSCore.
DOI 原文 · arXiv · PDF（开放获取） · DBLP
@inproceedings{HeJ0LS26,
title = {AGS:
<u>A</u>
ccelerating 3D
<u>G</u>
aussian Splatting
<u>S</u>
LAM via CODEC-Assisted Frame Covisibility Detection},
author = {Houshu He and Naifeng Jing and Li Jiang and Xiaoyao Liang and Zhuoran Song},
booktitle = {ASPLOS},
year = {2025},
doi = {10.1145/3760250.3762229}
}