search

acaglayan / exploitCNN-RNN

Exploiting Multi-Layer Features Using a CNN-RNN Approach for RGB-D Object Recognition (ECCV 2018 workshops)
12 stars 2 forks source link

Liblinear Test Accuracy: 0.000000 #1

Open itsagirl opened 5 years ago

itsagirl commented 5 years ago

When I run this demo by using the depth images from washington rgbd datasets, the accuracy is 0%. What can I do? Hope for your answer, plz.

Accuracy = 0% (0/692) Liblinear Test Accuracy: 0.000000

Classification result of Depth (Layer-3 + Layer-4)... Accuracy = 0% (0/692) Liblinear Test Accuracy: 0.000000

Classification result of Depth (Layer-3 + Layer-5)... Accuracy = 0% (0/692) Liblinear Test Accuracy: 0.000000

Classification result of Depth (Layer-4 + Layer-5)... Accuracy = 0% (0/692) Liblinear Test Accuracy: 0.000000

Classification result of Depth (Layer-3 + Layer-4 + Layer-5)... Accuracy = 0% (0/692) Liblinear Test Accuracy: 0.000000

acaglayan commented 4 years ago

Hi, sorry for such a late return. Hope, you solved the problem. I need to know what causes this, please run the code step-by-step, check if you have CNN features, then RNN features properly, and final phase is classification. For that, you also need to install liblinear properly.