Skip to main content
. 2025 Sep 17;11:e3080. doi: 10.7717/peerj-cs.3080

Table 3. Comparison of the three models with other standard models.

Model Inference time (ms) Params (M) Memory (MB) FLOPs MACs
Mask region-based CNN (Faster R-CNN) (Mask R-CNN) (Kriouile et al., 2024) 99 41 168 341 170
Faster region-based CNN (Faster R-CNN) (He et al., 2017) 82 31 121 230 115
Feature pyramid network (FPN) (Wang et al., 2025) 181 72 289 571 285
YOLOv8 (Chen et al., 2025) 15 7.8 31 17.1 77
VGG16 (Simonyan & Zisserman, 2014) 93 121 328 13.2 90
LSTM (Hochreiter & Schmidhuber, 1997) 12 1.9 7 1.4 13