search

opendilab / InterFuser

[CoRL 2022] InterFuser: Safety-Enhanced Autonomous Driving Using Interpretable Sensor Fusion Transformer
Apache License 2.0
548 stars 46 forks source link

weather used #45

Open a1wj1 opened 1 year ago

a1wj1 commented 1 year ago

In the paper, I did not see how many types of weather were used in the dataset. Is it the same as the code, using 14 types of weather?

a1wj1 commented 1 year ago

I have a few questions after reading the paper. Can you help me answer them? Thank you：

  1. Are the variables shown in the figure also manually set parameters? 1688178976889

  2. In the Limitation, you mentioned: Extending the current deterministic form to a probabilistic form can also help to deal with the uncertainty and multi-modality in detection and prediction. But I don't quite understand where the deterministic form in the method is reflected?

deepcs233 commented 1 year ago

Hi, about the weather , please refer to https://github.com/opendilab/InterFuser/issues/41#issuecomment-1614134553

Are the variables shown in the figure also manually set parameters?

Yes

But I don't quite understand where the deterministic form in the method is reflected?

InterFuser predicts other agents' status, future waypoints with a deterministic form. A deterministic form here means a discrete value, not a distribution (like gauss, laplace, etc. ).