search

nikhilroxtomar / FANet

FANet: A Feedback Attention Network for Improved Biomedical Image Segmentation [IEEE Transactions on Neural Networks and Learning Systems]
https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9741842
30 stars 8 forks source link

Do you add the data processed by data augmentation into the dataset? #3

Open lsh-med opened 1 year ago

lsh-med commented 1 year ago

I read your paper,The Table I show the number of data after data augmentation,but l don't see in the code.

tyjcbzd commented 9 months ago

Hello, I read your paper and code. I found that in your ResidualBlock. The SElayer in the structure diagram is behind the 3 x 3 convolution, but the SElayer in the actual code is behind the 1 x 1 convolution. Maybe this is a mistake? 66aa4f6acf094b210447889eec9db74

ruti0823 commented 5 months ago

Hello, I have also noticed your issue and improved the code according to the network structure diagram, but found that the performance is not as good as the original code. Have you solved this problem?

tyjcbzd commented 5 months ago

Hello, I have also noticed your issue and improved the code according to the network structure diagram, but found that the performance is not as good as the original code. Have you solved this problem?

No, maybe they use some tricks. The results can vary from devices to devices.

moonmisso commented 1 month ago

Hello, I have also noticed your issue and improved the code according to the network structure diagram, but found that the performance is not as good as the original code. Have you solved this problem?

Hello, may I ask what improvements you have made, and what is the final performance? Mine is ten points off