NPLSD：在NPU微控制器上加速线段检测
NPLSD: Accelerating Line-Segment Detection on NPU Microcontrollers
浏览论文内容
中文总结 AI 辅助
针对NPU微控制器上Transformer线段检测器不兼容的问题，提出NPLSD-H和NPLSD-M两种全卷积检测器，分别基于HGNetv2和M-LSD-tiny，在STM32N6上实现高精度检测，初始化贡献显著。
中文摘要 AI 辅助
线段检测是机器人技术、自主导航和工业检测的基础。虽然基于Transformer的检测器达到了最高的精度，但由于资源限制，它们在微控制器上的部署仍然不切实际。STM32N6凭借其Neural-ART NPU，有望在极边缘实现深度视觉。然而，现有的检测器依赖于注意力、网格采样和归一化，这些算子不受面向卷积的NPU支持。这种架构不匹配被逐算子地刻画：注意力、网格采样和归一化缺乏加速器原语，而解码器的自注意力单独就产生了39 MB的分数张量，超过了片上内存。为了解决这一限制，NPLSD被引入为一对基于一种设计方法论的NPU兼容线段检测器。NPLSD-H保留了LINEA的卷积HGNetv2骨干网络，并将Transformer头部替换为全卷积特征金字塔和F-Clip密集头部。NPLSD-M将M-LSD-tiny主干适配到支持的算子集。从ImageNet热启动并在ShanghaiTech Wireframe上训练，2.63M参数的NPLSD-H达到sAP^10=37.9（int8为35.9）；0.62M参数的NPLSD-M达到41.9（int8为41.1）。受控消融将主干隔离为唯一变量，仅初始化就贡献了4.6个点。
英文摘要
Line-segment detection is fundamental to robotics, autonomous navigation, and industrial inspection. While transformer-based detectors achieve the highest accuracy, their deployment on microcontrollers remains impractical due to resource constraints. The STM32N6, with its Neural-ART NPU, promises to enable deep vision at the extreme edge. However, existing detectors rely on attention, grid-sampling, and normalization, operators that are unsupported by the convolution-oriented NPU. This architectural mismatch is characterized operator by operator: attention, grid-sampling, and normalization lack accelerator primitives, and the decoder's self-attention alone materializes a 39 MB score tensor that exceeds on-chip memory. To address this limitation, NPLSD is introduced as a pair of NPU-compatible line-segment detectors built from one design methodology. NPLSD-H retains the convolutional HGNetv2 backbone of LINEA and replaces the transformer head with a fully-convolutional feature pyramid and an F-Clip dense head. NPLSD-M adapts the M-LSD-tiny trunk to the supported operator set. Warm-started from ImageNet and trained on ShanghaiTech Wireframe, the 2.63M-parameter NPLSD-H reaches sAP^10=37.9 (35.9 int8); the 0.62M-parameter NPLSD-M reaches 41.9 (41.1 int8). A controlled ablation isolates the trunk as the only variable, and initialization alone accounts for 4.6 points.
发表机构
- Iran University of Science and Technology（伊朗科技大学）
- Case Western Reserve University（凯斯西储大学）
机构由 AI 辅助整理，请以论文原文为准。
