Global metrics (TNR, Recall, Precision, and F1 score; see Table 2 for definitions) for the binary classification of interesting (with biosignatures/bioindicators) and non-interesting planets across different SNR values. The pink shaded area below 0.6 in the graphs indicates suboptimal classification performance. The solid lines represent the results for spectra without stellar contamination, while the coloured shaded areas around each curve indicate the variance in the results produced by stellar contamination.