Semantic segmentation of HeLa cells: An objective comparison between one traditional algorithm and four deep-learning architectures
Fig 9
Metrics used for comparison between image-processing algorithm and four pre-trained deep learning architectures—VGG16, ResNet18, Inception-ResNet-v2, and U-Net for semantic segmentation of HeLa cell imaged with Electron Microscopy (EM).
(Top row) Accuracy. (Middle row) Jaccard similarity index, also known as intersection over union, for all algorithms. Green box denotes the central slices and corresponding Jaccard similarity index that is magnified below. (Bottom row) Jaccard similarity index for central slices (slices between 75/300 and 225/300—interquartile range (IQR)), for easier comparison. The image-processing algorithm outperforms all deep learning architectures.