How to use the 4D model after the training has been completed?

[chenyuhan1997]

Thank you for open-sourcing your code, it's very inspiring. We currently want to be able to view the motion of a vehicle in a 3D scene. But it seems that this is not possible after training is complete. I also don't know how to change the perspective of the scene. It seems to be fixed. I hope you can answer my questions. Thanks

[LightwheelAI]

Hi, thanks for your trying! If you want to change the view, you could create a json with following format. Then you can modify the transform to control the view. After this step, you just need to use --vehicle-configof sgn-renderto specify this file, and you will get a novel view video.

[
    {
        "camera": "FRONT",
        "image_path_patten": ".*/FRONT/.*",
        "transform": [
            [
                1.0,
                0.0,
                0.0,
                0.0
            ],
            [
                0.0,
                1.0,
                0.0,
                0.0
            ],
            [
                0.0,
                0.0,
                1.0,
                0.0
            ],
            [
                0.0,
                0.0,
                0.0,
                1.0
            ]
        ]
    },

]