How to perform inference and evaluation using the provided BDD100K dataset

[HuBocheng]

Hi,

Thank you for releasing this great project!

I’m currently trying to run inference and compute evaluation metrics using the provided BDD100K dataset. But I noticed the dataset is stored in .h5 format and seems primarily structured for training purposes.

While looking through sample.py, I didn’t find any logic that samples from the BDD100K .h5 files or runs inference using them. I was wondering if you could clarify:
Is there an existing script or recommended way to run inference on the BDD100K dataset?

Thank you in advance!

[sirlancer]

Hi,
I have the same problem, do you solve it ?