Cheng-Lin Li comments

Results 6 comments of


                                            Cheng-Lin Li

Training Performance Do Not Improve

Hi AuliaRizky, You may need to try a single image to overfit the model first. If it works, then you know the model is complex enough for your task. Or...

Training Performance Do Not Improve

Hi AuliaRizky, 1. My implementation revises the loss_type from 'sorensen' to 'jaccard' in custom_losses.py, function: dice_soft. According to my previous discussion with the original author, he prefers 'sorensen' than 'jaccard'....

bad mask

Hi fenaux, Are you training for one single image then test the result? Or you already train the model in a human data set then test it? From the result,...

Training image and mask error

Please refer to the readme. 4-3-1. For 2D images Dimension of images: (Width, Height, Channels) Your input should be 3 channels (512, 512, 1)

Final output and raw output are always same no matter what input images are given

Hi, Please refer the readme 5.3 https://github.com/Cheng-Lin-Li/SegCaps#5-3-test-your-model The program will convert all image files into numpy format and store training/testing images into ./data/np_files and testing (and training) file lists under...

pdf

Hi zhiyang, Those two formulas are equivalent. Kr, Cheng-Lin Li On Thu, Sep 26, 2019 at 7:19 PM zhiyang wrote: > > https://github.com/Cheng-Lin-Li/MachineLearning/blob/5534263b4ad691e16d9d532ceeb94a1265ccd7a2/GMM/GMM.py#L81 > > should it be pow((2*math.pi), -d/2)...