Open comflife opened 1 year ago
No need to modify the code. Once the homography matrix is calculated, it doesn't need to be recalculated. However, each time, it is necessary to use the already computed matrix to perform a homography transformation on the new video frame.
Thank you for your kind response. I have one more question and would like to ask if you would be willing to provide a pre-trained model of your results.
Thank you for your kind response. I have one more question and would like to ask if you would be willing to provide a pre-trained model of your results.
The current focus of all authors has changed, and we are unable to provide it temporarily. Once we have the time, we will provide a retrained model, but it has already lost its timeliness. You can refer to the content in this link https://github.com/gigo-team/bev_lane_det/issues/10 , which has achieved metrics close to those in the paper.
@comflife Can you share the source code? Thanks,1017094591@qq.com
How can I use this with real time camera video? It is necessary to change codes in virtual camera?