Ma, Z.; Dai, P.; Zhuo, W.; Qin, X. One2Seq: One-Token Wise Decoder for Efficient Scene Text Recognition. AAAI 2026, 40, 7883-7891.