超轻量3D点云识别模型，参数少17倍仍保持高精度。
SLNet: A Super-Lightweight Geometry-Adaptive Network for 3D Point Cloud Recognition
- 用非参数自适应嵌入和几何调制单元构建轻量骨干
- 0.14M参数在ModelNet40达93.64%准确率，比同行少5倍参数
- 适合部署受限设备，兼顾精度与计算效率
我们提出SLNet，一种用于3D点云识别的轻量级主干网络，无需复杂注意力、图结构或深层MLP带来的高昂计算开销。模型基于两个核心思想：NAPE（非参数自适应点嵌入），通过高斯RBF与余弦基组合实现输入自适应带宽与融合以捕捉空间结构；以及GMU（几何调制单元），每通道仅引入2个可学习参数的仿射调制器。这些组件嵌入四阶段分层编码器中，结合FPS+kNN分组、非参数归一化与共享残差MLP。实验表明，极小模型仍具备强竞争力：在ModelNet40上，SLNet-S（0.14M参数，0.31 GFLOPs）达到93.64%整体准确率，优于参数多5倍的PointMLP-elite；SLNet-M（0.55M参数，1.22 GFLOPs）达93.92%，超越PointMLP但仅需其1/24参数。在ScanObjectNN上，SLNet-M达84.25%准确率，较PointMLP低1.2个百分点，却少28倍参数。针对大规模场景分割，SLNet-T引入局部Point Transformer注意力，在S3DIS Area 5上达58.2% mIoU，仅用2.5M参数，远低于Point Transformer V3的17倍以上。我们还提出NetScore+，在原基础上加入延迟与峰值内存评估，更贴近实际部署需求。多基准与硬件下，SLNet在精度与效率间取得良好平衡。代码已开源：https://github.com/m-saeid/SLNet。
原文摘要 · Abstract (English)
We present SLNet, a lightweight backbone for 3D point cloud recognition designed to achieve strong performance without the computational cost of many recent attention, graph, and deep MLP based models. The model is built on two simple ideas: NAPE (Nonparametric Adaptive Point Embedding), which captures spatial structure using a combination of Gaussian RBF and cosine bases with input adaptive bandwidth and blending, and GMU (Geometric Modulation Unit), a per channel affine modulator that adds only 2D learnable parameters. These components are used within a four stage hierarchical encoder with FPS+kNN grouping, nonparametric normalization, and shared residual MLPs. In experiments, SLNet shows that a very small model can still remain highly competitive across several 3D recognition tasks. On ModelNet40, SLNet-S with 0.14M parameters and 0.31 GFLOPs achieves 93.64% overall accuracy, outperforming PointMLP-elite with 5x fewer parameters, while SLNet-M with 0.55M parameters and 1.22 GFLOPs reaches 93.92%, exceeding PointMLP with 24x fewer parameters. On ScanObjectNN, SLNet-M achieves 84.25% overall accuracy within 1.2 percentage points of PointMLP while using 28x fewer parameters. For large scale scene segmentation, SLNet-T extends the backbone with local Point Transformer attention and reaches 58.2% mIoU on S3DIS Area 5 with only 2.5M parameters, more than 17x fewer than Point Transformer V3. We also introduce NetScore+, which extends NetScore by incorporating latency and peak memory so that efficiency can be evaluated in a more deployment oriented way. Across multiple benchmarks and hardware settings, SLNet delivers a strong overall balance between accuracy and efficiency. Code is available at: https://github.com/m-saeid/SLNet.
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。