Display-Semantic Transformer for Scene Text Recognition

• Published in: Sensors (Basel, Switzerland) Vol. 23; no. 19; p. 8159
• Main Authors: Yang, Xinqi, Silamu, Wushour, Xu, Miaomiao, Li, Yanbing
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 28.09.2023; MDPI
• Subjects: Accuracy; Computational linguistics; cross-modal attention; Efficiency; Language; Language processing; linguistic knowledge; Linguistics; Natural language interfaces; Neural networks; scene text recognition; Semantics; transformer; visual information; France; Germany

Linguistic knowledge helps a lot in scene text recognition by providing semantic information to refine the character sequence. The visual model only focuses on the visual texture of characters without actively learning linguistic information, which leads to poor model recognition rates in some noisy...