Table 5 Performance analysis of different protein particle types.
From: A hybrid YOLO-UNet3D framework for automated protein particle annotation in Cryo-ET images
Protein Particle | Recall | F4-score | IoU | AP | Localization Error | False Discovery Rate | Miss Rate |
|---|---|---|---|---|---|---|---|
apo-ferritin | 0.9885 | 0.9535 | 0.6091 | 0.6065 | 12.5889 | 0.3862 | 0.0111 |
beta-amylase | 0.6352 | 0.5816 | 0.2224 | 0.1702 | 17.6588 | 0.7514 | 0.3648 |
beta-galactosidase | 0.8681 | 0.6936 | 0.1609 | 0.1438 | 19.7642 | 0.8356 | 0.1319 |
ribosome | 0.9560 | 0.8879 | 0.4177 | 0.4075 | 24.8659 | 0.5748 | 0.0446 |
thyroglobulin | 0.8608 | 0.7333 | 0.2132 | 0.1906 | 24.9049 | 0.7794 | 0.1399 |
virus-like-particle | 1.0000 | 0.9314 | 0.4533 | 0.4533 | 19.4001 | 0.5467 | 0.0000 |
