Table 5. Summarises our model’s performance relative to recent literature.

Study year	Classifier	Accuracy (%)
2023	Hybrid RNN-LSTM	94.21
2023	Ensemble Deep-Learning Model	96.54
2024	Squeeze-and-Excitation ResNet	97.35
2025	Residual Attention BiLSTM	97.7
2025	X-TLRABiLSTM (Proposed)	98.2