Table 6 Comparison experiment results.

From: Linguistic-visual based multimodal Yi character recognition

Method

ACC /%

Training time/h

Year

CRNN15

0.939

28

2016

SRN26

0.948

42

2020

ViTSTR38

0.916

45

2021

RFL39

0.920

30

2021

ABINet28

0.961

48

2021

SVTR40

0.952

52

2022

Proposed

0.995

50

Â