free(): double free detected in tcache 2

HuXiaoling / TopoLoss

Code for the NeurIPS 2019 paper: Topology-Preserving Deep Image Segmentation

MIT License

146 stars 22 forks source link

free(): double free detected in tcache 2 #19

Closed mmmmimic closed 1 year ago

mmmmimic commented 1 year ago

Hi Xiaoling!

Thank you for the impressive work! I had some trouble when incorporating the TopoLoss into my model training. Specifically, I have two questions

Is the function getTopoLoss() intended for processing single image pairs? I would kindly like to know the shape of the two inputs, likelihood_tensor and gt_tensor, in PyTorch.
I met an error called "free(): double free detected in tcache 2" during training. Do you happen to have any idea about this error?

Thank you in advance!

HuXiaoling commented 1 year ago

@mmmmimic, thanks for your interest in our work! What do you mean by 'processing single image pairs'? The shapes of the two inputs are dependent on your data -- once the patch contains meaningful topological information.

Sorry that I haven't never seen such errors before.

mmmmimic commented 1 year ago

Thank you! Could the getTopoLoss() handle inputs in batch? If I am computing the TopoLoss for a multi-class segmentation task, would there be an efficient way to compute the loss for images in a batch?

HuXiaoling commented 1 year ago

The function getTopoLoss() was originally designed for single image pairs.

mmmmimic commented 1 year ago

Thank you!