RENO实现实时3D激光雷达点云压缩，速度达10帧/秒且模型仅1MB。
RENO: Real-Time Neural Compression for 3D LiDAR Point Clouds
- 跳过八叉树，直接用多尺度稀疏张量表示点云
- 一次完成占据编码，节省时间，实现10帧/秒实时压缩
- 模型仅1MB，比G-PCCv23和Draco分别省12.25%和48.34%码率
尽管基于学习的神经模型在激光雷达点云压缩（LPCC）任务中取得显著进展，但实现实时压缩——众多工业应用不可或缺的标准——仍是巨大挑战。本文提出RENO，首个面向3D激光雷达点云的实时神经编解码器，以轻量化模型实现优异性能。RENO跳过八叉树构建，直接基于多尺度稀疏张量表示。不同于多阶段推理，RENO设计稀疏占据编码，利用跨尺度相关性，一次性推导体素占据状态，大幅节省处理时间。实验表明，RENO在桌面平台（如单张RTX 3090 GPU）上实现14位深度下10帧/秒的编码与解码速度，同时在相近质量下，相比G-PCCv23和Draco分别降低12.25%和48.34%码率。模型大小仅1MB，适合实际部署。源代码已开源。
原文摘要 · Abstract (English)
Despite the substantial advancements demonstrated by learning-based neural models in the LiDAR Point Cloud Compression (LPCC) task, realizing real-time compression - an indispensable criterion for numerous industrial applications - remains a formidable challenge. This paper proposes RENO, the first real-time neural codec for 3D LiDAR point clouds, achieving superior performance with a lightweight model. RENO skips the octree construction and directly builds upon the multiscale sparse tensor representation. Instead of the multi-stage inferring, RENO devises sparse occupancy codes, which exploit cross-scale correlation and derive voxels' occupancy in a one-shot manner, greatly saving processing time. Experimental results demonstrate that the proposed RENO achieves real-time coding speed, 10 fps at 14-bit depth on a desktop platform (e.g., one RTX 3090 GPU) for both encoding and decoding processes, while providing 12.25% and 48.34% bit-rate savings compared to G-PCCv23 and Draco, respectively, at a similar quality. RENO model size is merely 1MB, making it attractive for practical applications. The source code is available at https://github.com/NJUVISION/RENO.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。