Table 1 Speckle-based DNN comparison. The table provides a detailed classification report evaluating the performance of the proposed DNN model for shape recognition tasks. The metrics include precision (the proportion of true positive predictions out of all positive predictions), recall (the proportion of true positive predictions out of all actual positives), confidence (the average of precision and recall, indicating overall classification reliability, higher is better), and interpretation (a qualitative assessment based on confidence and complexity, summarizing model performance per class. The results are presented for individual shape classes (e.g., Circle, Rectangle, Triangle), multi-shape versions of these shapes (e.g., multi Circle, multi Rectangle, multi Triangle), mixed shapes (Mix), and a baseline category (White and Black).

From: Visual cortex speckle imaging for shape recognition