Table 1 Model evaluation metrics.
From: YOLO-Granada: a lightweight attentioned Yolo for pomegranates fruit detection
Items | Formulas |
|---|---|
Detection accuracy | |
P | \(Precision=TP-(TP+FP)\) |
R | \(Recall=TP/(TP+FN)\) |
F1 | \(F1 scores=(2 \times P \times R)/(P+R)\) |
mAP@0.5 | \(mAP=\left( \sum _{i=1}^{k} AP_i\right) /k\) |
Model complexity | |
Size | Modelweightsize |
Params | \(Params=\sum (K_h \times K_w \times C_{in} \times C_{out})\) |
FLOPs | \(FLOPs=\sum (K_h \times K_w \times C_{in} \times C_{out} \times H \times W)\) |
Detection speed | |
FPS | \(FPS=1000/(pretreated+inference+NMS)\) |
