SpringerPlus

Table 6 A comparison of the presented system on UPTI dataset with other techniques reported in the literature

From: Urdu Nasta’liq text recognition using implicit segmentation based on multi-dimensional long short term memory neural networks

Authors	Features	Approach	UPTI dataset	Ave. char. accuracy (%)
Ul-Hasan et al. (2013)	Pixels	BLSTM	46% train set	94.85
			34% validation set
			20% test set
Ahmed et al. (2016)	Pixels	BLSTM	46% train set	88.94
			44% validation set
			10% test set
Naz et al. (2015a)	Statistical features	MDLSTM	68% train set	94.97
			16% validation set
			16% test set
Proposed system	Pixels	MDLSTM	68% train set	98 ± 0.25
			16% validation set
			16% test set

Back to article page