← 返回 JSSC 论文列表JSSC 2022第2期Digital Circuits
A Pipelined Point Cloud Based Neural Network Processor for 3-D Vision With Large
提出一种基于点云的流水线神经网络处理器,用于移动设备中的3D视觉系统,优化了采样分组层和卷积层的加速。
4.45 ms处理时间,8.24 mm HPE误差,266 mW功耗
点云神经网络处理器3D视觉移动设备流水线加速
▸创新点1:流水线异构架构加速PNN - 该处理器采用流水线异构架构,专门针对点云神经网络(PNN)中的采样分组层(SGLs)和卷积层(CLs)进行优化,通过并行处理和数据流优化,实现了低延迟的3-D智能系统,处理时间仅为4.45 ms,功耗为266 mW。
▸创新点2:基于窗口的采样分组算法提升吞吐量 - 通过窗口采样分组(WSG)算法,直接从深度图像中对3-D点云数据进行采样和分组,避免了传统方法的冗余计算,将SGLs的吞吐量提升了2.34倍,显著提高了处理效率。
▸创新点3:最大池化预测核心减少延迟 - 设计了最大池化预测核心(MPPC),用于预测大规模(64-和128-to-1)的最大池化层,通过提前计算和预测,将吞吐量提升了1.31倍,同时解决了卷积核心(CC)中内存的bank冲突问题。
▸创新点4:分块数据预测隐藏延迟 - 通过对分块数据进行最大池化预测,有效隐藏了MPPC的延迟,进一步优化了系统的整体性能,使其在移动设备中实现了高效的3-D视觉处理。
Abstract
The point cloud data provides useful geometric
information to 3-D intelligent systems such as autonomous
driving, 3-D reconstruction, and hand pose estimation (HPE).
Many mobile devices have implemented the 3-D intelligent
system with their limited hardware resources. However, previous
processors were not designed for accelerating the point cloud
based neural network (PNN) which consists of sampling-grouping
layers (SGLs) and convolution layers (CLs). In this article,
a pipelined PNN processor i