search

charlesq34 / pointnet

PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
Other
4.73k stars 1.45k forks source link

some general questions #70

Closed moeinp70 closed 6 years ago

moeinp70 commented 6 years ago

Hi. I've read your paper (point net) and I really appreciate your work. I have some general questions about your work as follow :

  1. Is the reported test accuracy in the paper highest accuracy achieved? Or its calculated using other approaches, maybe averaging test accuracy over multiple times? . The way that test accuracy is calculated could be really useful for me.

  2. Is the value of num_votes always equal to one in the evaluation script?

3.What are the criteria for saving the trained model ? Under which criteria is the "log/model.ckpt" saved?

Thanks

charlesq34 commented 6 years ago

Hi @moeinp70 Thank you for your interest!

For the first two questions please see my answers here: https://github.com/charlesq34/pointnet/issues/49

As for 3: I'm not fully sure what you means but how the model is saved is specified by Tensorflow which helps save both the parameters of the trained network as well as training dynamics like iteration number, learning rate etc.

Hope it helps. Best, Charles