Camera-Based Monocular Depth Estimation in Orthodontics: Vision Transformer vs. CNN Model Performance.

However, this difference did not reach statistical significance (χ 2 = 1.19, p = 0.276), suggesting that although a trend toward reduced performance in high-complexity cases was observed, the model maintained robust accuracy across varying levels of task difficulty.

also in 111,027 other papers

Pairwise Model Comparisons Pairwise McNemar tests revealed highly significant differences in ranking performance among the models ( Table 3 ).

also in 132,142 other papers