William Ljungbergh

Hi, Thank you for a great dataset and development kit. I was wondering if you could shed some light on what coordinate systems are used throughout the code. Especially the...

Results using ST-P3 and VAD metrics

10

Hi, As mentioned [here](https://github.com/OpenDriveLab/UniAD/issues/29#issuecomment-1717594344) you are not using the same definition of the L2 metric as other works (e.g., [VAD](https://github.com/hustvl/VAD) or [ST-P3](https://github.com/OpenDriveLab/ST-P3)). However, you present UniADs numbers using your new...

As described in #166, there is a discrepancy between how planning metrics (L2 and collision rate) are computed in this repo vs previous works, such as [VAD](https://github.com/hustvl/VAD) and [ST-P3](https://github.com/OpenDriveLab/ST-P3). However,...

Question about L2 computation

2

Hi, and thank you for your work. When reviewing your evaluation code I find that your computation of the L2 displacement error ([here](https://github.com/hustvl/VAD/blob/1ede800a517653837faf78b5ec1c3b5fc3ec77b5/projects/mmdet3d_plugin/VAD/VAD.py#L624C1-L629C1)) is computed using the average displacement error...

Training on different datasets

1

Hi, Thank you for your work. I was wondering if you are aware of any efforts in training/validating your work on other datasets e.g., Argoverse2, Waymo, or ZOD?