New Trajectories Generation
Hello,
Thank you for your excellent work! I have a question regarding modifying val_data to generate new trajectories. Specifically, I would like to simulate a lane change by shifting the camera to the left and then generate new videos or images.
Could you suggest a simple way to achieve this? What modifications would be necessary in val_data, such as in val_data["camera_parameter"]?
Thanks in advance!
i tested add offset to camera_param before enter into pipe,but the generated images are still original view point
Please consider building them like real trajectories such as those from the validation set. It may not work on arbitrary parameters.
Please consider building them like real trajectories such as those from the validation set. It may not work on arbitrary parameters.
Thanks for your suggestion! i tried modified the cam pose and transform the bbox in val_data, In some scenarios, a small offset can work, but a large offset may cause problems in the model's cross-view and scene structure generation.
Like I said, please do your best to mimic a real trajectory and see what will happen. I think you should be able to find real trajectory as examples for turning, speed changing, lane changing in the dataset.
Please understand I can only offer help to basic usage of our models, and may not have time to look into details of your use cases.
Since this is an open sourced work, it would be helpful to share details of your experiences/issues/solutions. Others may reuse your experience or shed some light on it.
This issue is stale because it has been open for 7 days with no activity. If you do not have any follow-ups, the issue will be closed soon.