Table 4.

The AUC and F-score results of the proposed approach and the existing approaches.

Author	Method	AUC (%)	F-Score (%)
Islam et al. [40]	CNN	80.50	85.00
Jordi et al. [41]	VGG16	88.71	81.76
Li et al. [42]	ResNet101	93.00	91.30
Wang et al. [44]	EffifinetB3	73.00	89.00
He et al. [43]	ResNet models	92.70	90.70
Gour and Khanna [45]	Two I/P VGG16	84.93	85.57
Proposed Approach	(R-CNN+LSTM)+NCAR+SVM	97.00	89.97