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

Figure 5

Structure of the attention-based decoder in the recognition branch.