Furthermore, we implement some of the widely used loss functions such as cross entropy (CE), Dice loss (DL) and Tversky loss (TL) on a noise-robust intermediate false positive鈥揻alse negative matrix model to evaluate the performances of these loss functions when dealing with noisy data....
In literature, most research work uses smoothing for handling noise in the segmentation, which results in the loss of fine structure of images. Moreover, the non-linearity present in the data is also not addressed well, resulting in the loss of crucial features. In this research work, we ...