search

OpenDriveLab / Openpilot-Deepdive

Our insights of Openpilot, a deepdive project on it
MIT License
231 stars 66 forks source link

some questiones on nuscene #16

Closed promesse closed 1 year ago

promesse commented 1 year ago

Thanks for your great work! However i meet a problem when i train and demo for nuscene I had already follow your issue(https://github.com/OpenPerceptionX/Openpilot-Deepdive/issues/9) Maybe because i just train for 10epoch in nuscene mini batch, when i demo it, the prediction is all equal to 0 Have you ever meet it? And i changed the ground truth to draw on the pic, the direction also seem not right Car was clearly going to the left image

promesse commented 1 year ago

meanwhile when you use gru to predict the 10 waypoints, why not use grucell to predict one by one? does it make sense?

image

ElectronicElephant commented 1 year ago

And i changed the ground truth to draw on the pic, the direction also seem not right Car was clearly going to the left

You should double check your coordinate system. It's the fundamental requirement before you do anything in the real world.

promesse commented 1 year ago

?????? I use your demo code and find it

promesse commented 1 year ago

why you close it?

promesse commented 1 year ago

And i changed the ground truth to draw on the pic, the direction also seem not right Car was clearly going to the left

You should double check your coordinate system. It's the fundamental requirement before you do anything in the real world.

-_-''

ElectronicElephant commented 1 year ago

As is shown in the official website (https://www.nuscenes.org/nuscenes#data-collection), the coordinate system used by the IMU in nuScenes dataset is not the same as the one we adopt when we implementing the model on Comma2k19 dataset.

I think you can simply flip the y axis and you will get the correct result.