zexijia/evsegformer

evsegformer's Issues

MVSEC Dataset for Semantic Segmentation

Hi @zexiJia @hewh16, thanks for open-sourcing this work.

Regarding the mentioned experiments on the MVSEC dataset, where can I download the data and associated semantic labels? Thanks!

关于生成MVSEC label的链接失效

Dataset Download Link is Broken

Thank you for sharing this project. While going through your project, I noticed that the download link provided for the ground truth of dataset ddd17 is no longer working. Would you be able to provide an updated download link, so we can continue using the dataset in your project?

Here is the broken download link: https://drive.google.com/open?id=1Ug6iZc7WYQWCklxwcemCeyw3CPyuuxJf

Looking forward to your response, thank you very much!

Best regards

About EvSegFormer 's 6 frame input data

Hello,
I am someone interested in your research. I am trying to convert the h5 event data of DSEC semantic into your numpy array format based on the formulas in your paper, and I have a few questions. I would be really grateful if you could answer them.

I am curious about how negative values exist in the array.
There are many -1 values in the first and second frames of your array (presumed to be areas where no events occurred). I am curious about how these -1 values are generated.
I am curious about how the six frames are stacked in your array. I assume they are stacked as S(-1), S(+1), M(-1), M(+1), Hist(-1), Hist(+1).

I would be very happy if you could answer