Tommy Hu

icon indicating an email [email protected]

icon location Shenzhen, China icon location Focus on Autonomous Driving and Robotics. My journey starts from Visual Perception, SLAM and Embedded system, and extend to Motion Prediction and VLM/VLA.

Results 5 issues of Tommy Hu

Missing the camera to lidar extrinsic in 3dbox_to_cloud.py

There are two steps transformation in the 3dbox_to_cloud.py: 1. object coordinate to camera coordinate 2. camera coordinate to lidar coordinate The bug exists in the second step, it doesn't consider...

How to visualize the infer result?

Great work and thanks for sharing! I have run the infer.py successfully, but the output is plain text, is there a out-of-the-box visualization tool that I can show the pointcloud...

CenterPoint Training Error: RuntimeError: (PreconditionNotMet) The meta data must be valid when call the mutable data function.

4 comment

The error is reported from [apollo-model-centerpoint](https://github.com/ApolloAuto/apollo-model-centerpoint), which uses paddle as backend, I found a similar issue here, but looks like not the same root cause: https://github.com/PaddlePaddle/Paddle3D/issues/118 below is my error...

Abnormal demo.py result using nuscenses dataset pre-trained models

1 comment

This issue is related to [issue 1341](https://github.com/open-mmlab/OpenPCDet/issues/1341), but here we track a different problem. After modify the demo.py a bit in order to run the nuscenses dataset pre-trained model, I...

Not quite understanding the occupancy along the height

Question about the FlashOCC paper: After extracting the image 2D feature, most of the subsequent modules, except the final Channel2Height operator, operate on the 2D BEV feature, how does the...