Sensitivity and Specificity are performance metrics used to evaluate classification models (especially in imbalanced or medical diagnostics contexts). Sensitivity, also known as recall or true positive rate, is defined as the proportion of actual positive instances which are correctly identified as positive​In formula: Sensitivity = TP / (TP + FN), where TP = true positives and FN = false negatives. A model with high sensitivity has few false negatives (it misses few true positives). Specificity (the true negative rate) is the proportion of actual negative instances correctly identified as negative​In formula: Specificity = TN / (TN + FP), where TN = true negatives and FP = false positives. High specificity means the model commits few false alarms. Together, sensitivity and specificity characterize a classifier’s ability to detect positives and negatives; in practice there is often a trade-off between them (e.g., via the decision threshold) depending on whether false negatives or false positives are more costly.Users can prompt SAM with various inputs (clicks, bounding boxes, or text prompts) and the model will return the corresponding segmentation mask. This flexibility lets SAM “segment anything” new, adapting to new tasks or domains with impressive out-of-the-box performance and without the need for additional training.