search

hossein1387 / U-Net-Fixed-Point-Quantization-for-Medical-Image-Segmentation

Repository containing code for "U-Net Fixed-Point Quantization for Medical Image Segmentation" paper at MICCAI2019
MIT License
83 stars 17 forks source link

How to use the inference function to get more output images? #5

Closed ShaoCorn closed 3 years ago

ShaoCorn commented 4 years ago

Hi, I have tried your great network and it trained well. But when I was using the "inference" function, I don't know how to utilize the trained model to inference more images, and it seems that when I was using "inference" mode, the network started to retrain the model (emdataset). Thanks!

1gzcdestiny1 commented 4 years ago

hi。I appreiate the code you provided.But i couln't find the code about showing the the network size,can you tell me what the method you did to show the size of the network after quantization?

hossein1387 commented 3 years ago

Hi @ShaoCorn , could you please clarify which network and which quantization you are referring to? Also, could you please send me the command you are using for training? Have you changed any part of this repo locally that might break the training process?

hossein1387 commented 3 years ago

@1gzcdestiny1 Please refer to https://github.com/hossein1387/U-Net-Fixed-Point-Quantization-for-Medical-Image-Segmentation/issues/6