Change camera pose to identity matrix (np.eye(4)) and move object to get same result

[dqj5182]

Hi!

First of all, thanks for your amazing rendering framework!
Pyrender helped my colleagues and me a lot on rendering various 3D models in python.

My situation is that I have focal length, principal point, camera pose, and a certain mesh.
Based on these information, I render with "pyrender.OffscreenRenderer" with "focal length and principal point" as inputs for "pyrender.IntrinsicsCamera" to initialize camera and add camera pose by "scene.add(camera, pose=cam_pose)" where "cam_pose" denote my camera pose information. And of course I add my mesh information to "scene.add(mesh, 'mesh')".

But I wish to change the camera pose information (which is usually added in "scene.add(camera, pose=cam_pose)") to identity matrix (np.eye(4)) and move by object mesh vertices so that it can render properly.

May I ask how I should move my object mesh vertices in order to make the rendering work?

[dqj5182]

I have tried to make the cam_pose as homogeneous matrix and make inverse of the matrix with "np.linalg.inv" to multiply the inverse matrix to object mesh vertices but it does not work.