lift-splat-shoot

lift-splat-shoot copied to clipboard

How to obtain the context vector c , cd, and distribution

1

I have read the paper carefully, and found no detailed information on how to generate those vectors, any one can explained them at length? and what's the distribution form? is...

hitbuyi

This PR adds the following changes: 1. A handful of changes in the model method to accomodate utilization of LIDAR data 2. Take the sumproduct of depth and the alpha...

adiser

Why still cropping imgs during validation phase?

1

Hello, thanks for your excellent work~ I find that during validation phase the data process still crop imgs, and this would decrease the fov of imgs, so why still do...

Zhenghao97

1

WH1365919

How to use the bev radar of nuscenes data and encode it instead of the lidar top data

Unicorncosmos

Code and dataset for map segmentation task

1

Hi! Thank you for your wonderful work. Currently, the repo only has weights and code for the detection task. How can we run the repo to reproduce the map segmentation...

hocherie

Hello and thank you for your outstanding work. I would like to ask why the rank values are calculated in this way, and does the expression ensure that the rank...

ZMC12342

How can I train on my own dataset? My dataset has only one camera, how do I build data.py?

1

I want to train on my own dataset, the [V2X dataset](https://github.com/AIR-THU/DAIR-V2X#dataset) In V2X-V dataset, there is only one camera, how do I build the data.py file?

Camellia-hz

why “get_binimg” Donot crop and resize

1

Randomly crop and resize image data， why get_binimg Donot crop and resize https://github.com/nv-tlabs/lift-splat-shoot/blob/d74598cb51101e2143097ab270726a561f81f8fd/src/data.py#L171C19-L171C19 ``` def get_binimg(self, rec): egopose = self.nusc.get('ego_pose', self.nusc.get('sample_data', rec['data']['LIDAR_TOP'])['ego_pose_token']) trans = -np.array(egopose['translation']) rot = Quaternion(egopose['rotation']).inverse img =...

wuzuowuyou

Communicate on WeChat groups， 微信群交流自动驾驶问题

1

You can add me to the question about autonomous driving, and I will bring you to the WeChat group for communication。 vx：837007389 备注自动驾驶 微信群交流自动驾驶问题，

wuzuowuyou