Hello, I specialize in multi-object tracking. In sparse4d_v3, I understand that there seems to be no need for modifications during the training phase, and that ID assignment is only necessary during the inference stage. I want to know where this is reflected, and how I should visualize the tracking results, like a video sequence with 3D boxes labeled with IDs.
Hello, I specialize in multi-object tracking. In sparse4d_v3, I understand that there seems to be no need for modifications during the training phase, and that ID assignment is only necessary during the inference stage. I want to know where this is reflected, and how I should visualize the tracking results, like a video sequence with 3D boxes labeled with IDs.