Giters

MIC-DKFZ / nnDetection

nnDetection is a self-configuring framework for 3D (volumetric) medical object detection which can be applied to new data sets without manual intervention. It includes guides for 12 data sets that were used to develop and evaluate the performance of the proposed method.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Evaluation on independent test set

dzeego opened this issue · comments

commented

❓ Question

Dear @mibaumgartner,

I am trying to evaluate the predictions of a pretrained model on an independent test set to which I have the labels.
When launching nndet_eval with the --test argument I get an error that the .npz files of the labels are not found in preprocessed/labelsTs.
It seems to me that I have to preprocess the data in labelsTs.
So, how can I perform such an evaluation?

Many thanks in advance

commented

Hi, the labels need to be placed in raw_splitted/labelsTs und you need to rerun the preprocessing -> it will skip most of the actual preprocessing and only convert the labels from the nifty images into actual bounding box coordinates (located in the npz files) which are used for the evaluation.

commented

This issue is stale because it has been open for 30 days with no activity.

commented

This issue was closed because it has been inactive for 14 days since being marked as stale.