.png){kind=icon}
.png){kind=icon}
Return
A Low-Latency Framework With Algorithm-Hardware Co-Optimization for 3-D Point Cloud
.png){kind=icon}
PRE
AI
DOI：10.1109/TCSII.2023.3283142.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
An an important type of 3D representation, the point cloud is widely used in many applications, such as autonomous driving, AR/VR, and intelligent robots, which require real-time interactions with humans. However, the sparsity of 3D point cloud data leads to severe computational inefficiency when being processed by 2D data processors, posing a huge challenge for hardware acceleration. In this brief, we aim at solving the inefficiency problem by algorithm-hardware co-optimization. Firstly, a lightweight network, named LPN, is proposed for point cloud data classification, which is 30x smaller than pointnet and still has comparable accuracy. Secondly, a reconfigurable computing core, named RCC, together with an adaptive dataflow, is developed to support different layers of the LPN. Specifically, to accelerate memory-intensive layers, a partially-parallel computing scheme is introduced to minimize the on-chip memory requirements and DRAM accesses. Finally, based on the above innovations, a low-latency accelerator is proposed to realize real-time computation for the point cloud, which is implemented on the Xilinx Kintex UltraScale KCU150 FPGA board. Experimental results show that it achieves 1.5x throughput improvement compared with the state-of-the-art works, and 35x speedup over Intel Xeon Gold 6148 CPU, demonstrating the superiority of the proposed method. The code of LPN is available from https://github.com/snowsil/LPN-model-for-3D-classifification.
Keywords:
Point cloud
lightweight network
hardware design
low-latency
FPGA
Journal
I
IF:
4.9
Papers:
8.8K
Citations:
2.5W