princeton-computational-imaging / nlosfeatureembeddings Goto Github PK

Thank you for uploading the relevant information of great work in detail.

Do you have any plans to release other dataset besides bike?

Hi, thanks for the great work!
I want to use the code you provided for rendering to generate synthetic transient data. After configuring the environment according to the guidelines of the md file, it still cannot run. Is there some cmake file that is not provided? I would like to ask how to run these codes to generate synthetic transient data.

Hi there!

I've tried to train/evaluate the model defined in deepVoxel.py on my own data, but it either fails assert statements or concat calls within the lct/fk/phasor layers whenever the batch size is bigger than 1. The same happens when running the standalone examples of these modules, provided in "utils_pytorch/". Is there a hack on how to reshape the inputs in order to process large batches? Or is the implementation inherently limited to process inputs one-by-one?

Thank you!
Best

Dear Wenzheng Chen,
Dear Fangyin Wei,

thank you for making your excellent work publicly available!

As there are just a few comments throughout the code, I wonder if you could help me to clarify couple of operations within the network class "DeepVoxels", defined in "DL_inference/network7_256/deepvoxel.py". Between lines 110-115 the volume x, which has size (256,128,128) after applying LCT, is then resized/cropped along the zdim=256 by doing zdimnew = zdim * 100 // 128, which yields a zdimnew=200. Is there a particular reason as for why the volume should be resized at this point? Also, why 100 and 128? Is there a way to avoid hardcoded numbers in that step?

Best regards

Thanks for the great work!

Would you also release the code for modeling detector noise (eq. 4 in the paper)? I am a little confused by how exactly time jitter is modeled. How many samples did you draw from the jitter distribution for each temporal bin? Thanks!