New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
converting .mat model to struct failed #2
Comments
i also meet the same question, do you solve it? |
I forgot if I resolve it eventually, maybe it was caused by the environment configuration of Matlab-NetVLAD. But after that, I turned to the Pytorch version, available at https://github.com/Nanne/pytorch-NetVlad |
I also work on pytorch version of netvlad. So did you figure out how to use vd16_pitts30k_conv5_3_vlad_preL2_intra_white.mat in pytorch? |
The trained model on Matlab cannot be used in PyTorch. Actually, PyTorch-version NetVLAD has its own trained checkpoints, available at the link above. |
Do you think this checkpoint as begin of training can work well with grey-scale image as input. BTW, do you read the paper 'Kidnapped radar' before? |
I got an error when converting model trained by Matlab vd16_pitts30k_conv5_3_vlad_preL2_intra_white.mat into struct using matlab/net_class2struct.
The error says that "Index exceeds matrix dimensions". And I found that the net.layers imported from vd16_pitts30k_conv5_3_vlad_preL2_intra_white.mat had empty members, which caused the error.
The same problem occurs in the AlexNet models trained by myself on Matlab.
How can I solve this problem?
Thx.
The text was updated successfully, but these errors were encountered: