search

kaixin96 / PANet

Code for our ICCV 2019 paper PANet: Few-Shot Image Semantic Segmentation with Prototype Alignment
322 stars 64 forks source link

Can't train the accuracy of the paper #17

Closed LiGa0507 closed 4 years ago

LiGa0507 commented 4 years ago

First of all, thank you very much for your open source code. And I have a question for you. The mean-IoU of 1-way 1-shot segmentation on PASCAL-5i dataset in your paper is 48.1 , but the result of my training with your code is 41.9 .Do you have any suggestions or tricks for me? Look forward to your reply.

kaixin96 commented 4 years ago

Hi @LiGa0507 , can you check which split does not match the reported performance? Thank you.

LiGa0507 commented 4 years ago

I see where the problem is, thank you!

tymatfd commented 4 years ago

I cannot get the reported result too. Here's my final result

INFO - main - ----- Final Result ----- INFO - main - classIoU mean: [0.56912405 0.50470607 0.50892423 0.32237848 0.16499783] INFO - main - classIoU std: [0.02153639 0.02453332 0.02968146 0.01232205 0.02166889] INFO - main - meanIoU mean: 0.4140261347733086 INFO - main - meanIoU std: 0.007181320264279855 INFO - main - classIoU_binary mean: [0.88718143 0.43176332] INFO - main - classIoU_binary std: [0.00223838 0.00538581] INFO - main - meanIoU_binary mean: 0.6594723762116546 INFO - main - meanIoU_binary std: 0.0026686067206232217 INFO - PANet - Completed after 0:26:07

Model is trained under default settings. How did you problem fixed? @LiGa0507, thanks for reply

CLCxxx commented 3 years ago

I also have this question 。could you tell me how did you fix this problem when you are free？thanks

CLCxxx commented 3 years ago

I find the problem，thx