Heterogeneous acceleration system and method based on CTPN network

A technology for accelerating systems and networks, applied in biological neural network models, architectures with a single central processing unit, neural architectures, etc. Effect
CN112732638AActive Publication Date: 2021-04-30SHANGHAI JIAO TONG UNIV +1
9 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN ยท China
Current Assignee / Owner
SHANGHAI JIAO TONG UNIV
Publication Date
2021-04-30

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention provides a heterogeneous acceleration system and method based on a CTPN network. The heterogeneous acceleration system comprises a CPU end and an FPGA end. the FPGA end comprises a first sub-graph and a second sub-graph, and the CPU end comprises a third sub-graph; the first sub-graph comprises a CTPN network CNN part, the second sub-graph comprises an RNN part, and the third sub-graph comprises a CTPN network remaining part; the first sub-graph and the second sub-graph are executed at an FPGA end, and the third sub-graph is executed at a CPU end; the output of the FPGA end is used as the input of the third sub-graph; and the CPU end finally realizes network inference and obtains a final result. According to the system and method, the deduction speed of the CTPN network can be greatly improved under the condition that the precision is slightly reduced, so that the accelerator can better realize a real-time scene character recognition function.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The present invention relates to the field of neural network accelerators, in particular to a CTPN network-based heterogeneous acceleration system and method. Background technique

[0002] Connectionist Text Proposal Network (CTPN), a neural network proposed in 2016 for scene text recognition, is actually a neural network based on CNN+RNN structure. Among them, CNN is used to extract deep features, and RNN is used for character sequence feature recognition. This network takes advantage of the respective advantages of CNN and RNN. RNN also uses bidirectional LSTM (BiLSTM) to recognize and infer text from different directions of the picture, which greatly improves the accuracy of text detection. Since various applications of scene text recognition require network recognition and inference to be as fast as possible, it is particularly important to implement hardware acceleration for CTPN networks. However, its complex network structure makes hardware ac...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More