Towards Accurate Scene Text Recognition with Semantic Reasoning Networks
This work improves scene text recognition accuracy and efficiency for applications like document analysis and image understanding, representing a novel method for a known bottleneck rather than an incremental advance.
The paper tackles the problem of scene text recognition by addressing the limitations of RNN-based methods, such as time-dependent decoding and one-way semantic context transmission, and proposes a semantic reasoning network (SRN) with a global semantic reasoning module (GSRM) that achieves state-of-the-art results on 7 benchmarks and offers significant speed advantages.
Scene text image contains two levels of contents: visual texture and semantic information. Although the previous scene text recognition methods have made great progress over the past few years, the research on mining semantic information to assist text recognition attracts less attention, only RNN-like structures are explored to implicitly model semantic information. However, we observe that RNN based methods have some obvious shortcomings, such as time-dependent decoding manner and one-way serial transmission of semantic context, which greatly limit the help of semantic information and the computation efficiency. To mitigate these limitations, we propose a novel end-to-end trainable framework named semantic reasoning network (SRN) for accurate scene text recognition, where a global semantic reasoning module (GSRM) is introduced to capture global semantic context through multi-way parallel transmission. The state-of-the-art results on 7 public benchmarks, including regular text, irregular text and non-Latin long text, verify the effectiveness and robustness of the proposed method. In addition, the speed of SRN has significant advantages over the RNN based methods, demonstrating its value in practical use.