HRNet Encoder and Dual-Branch Decoder Framework-Based Scene Text Recognition Model

Figure 4

Structure of multiscale fusion in the recognition branch.