Camera Position Optimization with phong_renderer about pytorch3d HOT 4 CLOSED

raster_settings = RasterizationSettings(
    image_size=256, 
    blur_radius=np.log(1. / 1e-4 - 1.) * blend_params.sigma, 
    faces_per_pixel=100, 
    bin_size=0
)

class SoftPhongShader(nn.Module):
    def __init__(
        self,
        device="cpu",
        cameras=None,
        lights=None,
        materials=None,
        blend_params=None,
    ):
        super().__init__()
        self.lights = (
            lights if lights is not None else PointLights(device=device)
        )
        self.materials = (
            materials if materials is not None else Materials(device=device)
        )
        self.cameras = (
            cameras
            if cameras is not None
            else OpenGLPerspectiveCameras(device=device)
        )
        self.blend_params = (
            blend_params if blend_params is not None else BlendParams()
        )

    def forward(self, fragments, meshes, **kwargs) -> torch.Tensor:
        texels = interpolate_vertex_colors(fragments, meshes)
        cameras = kwargs.get("cameras", self.cameras)
        lights = kwargs.get("lights", self.lights)
        materials = kwargs.get("materials", self.materials)
        colors = phong_shading(
            meshes=meshes,
            fragments=fragments,
            texels=texels,
            lights=lights,
            cameras=cameras,
            materials=materials,
        )
        images = softmax_rgb_blend(colors, fragments, self.blend_params)
        return images

phong_renderer = MeshRenderer(
    rasterizer=MeshRasterizer(
        cameras=cameras, 
        raster_settings=raster_settings
    ),
    shader= SoftPhongShader(device=device, lights=lights, blend_params=blend_params)
)