High Throughput Hardware/Software Heterogeneous System for RRPN-Based Scene Text Detection

Rotation Region Proposal Networks (RRPN) are used to generate rotated proposals with the information of text angle for arbitrary oriented scene text detection (STD). However, the computational complexity of RRPN inference is relatively high compared with other methods, which makes it difficult for m...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on computers 2022-07, Vol.71 (7), p.1507-1521
Main Authors: Xin, Yao, Chen, Donglong, Zeng, Chongyang, Zhang, Weichen, Wang, Yi, Cheung, Ray C. C.
Format: Article
Language:English
Subjects:
Algorithms
Central Processing Unit
Central processing units
Complexity
Computer architecture
Convolutional neural networks (CNN)
CPUs
Field programmable gate arrays
filed programmable gate array (FPGA)
Floating point arithmetic
Graphics processing units
Hardware
hardware architecture for machine learning
Proposals
Rotation
rotation region proposal networks (RRPN)
scene text detection (STD)
Software
Systems design
Throughput
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Rotation Region Proposal Networks (RRPN) are used to generate rotated proposals with the information of text angle for arbitrary oriented scene text detection (STD). However, the computational complexity of RRPN inference is relatively high compared with other methods, which makes it difficult for massive deployment. In this paper, the first full-stack FPGA-CPU heterogeneous system design of RRPN-based STD algorithm is proposed. A hardware/software partition method is presented to analyze and split the tasks to enhance the computation efficiency of hardware. The fast 2D Winograd algorithm and block floating point are utilized to reduce computation complexity while maintaining a relatively high precision. The implementation results show that the peak performance of MAC arrays in the proposed architecture reaches 655.4 GOPS and the energy efficiency achieves 64.9 GOPS/W. By fully exploiting the parallel and pipelined merits in the algorithms, the first hardware architectures for skew non-maximum suppression (S-NMS) layer and rotation region-of-interest (RRoI) polling layer are proposed. The throughput of the proposed hardware/software heterogeneous system achieves 40 times and 1.4 times improvements compared with CPU and GPU, respectively. Moreover, the comprehensive operating expense ratio of pure CPU, GPU, and the proposed system is 80.7:2.5:1, which indicates that it is suitable for massive deployment.
ISSN:0018-9340
1557-9956
DOI:10.1109/TC.2021.3092195