Posture generation color contrast

[wwwpkol]

Thank you for such an experienced work, the visualization has tortured me to the point of unconsciousness!

I am currently working on joint prediction. I have seen comparisons between other people’s work and GT. The greater the difference, the more obvious the color. Do you have any examples of this? Thank you so much!

[kaufManu]

Yes, this can be achieved with some color maps from matplotlib (not a dependency of aitviewer, so to be installed if it's not yet in your environment):

import os
import numpy as np

from aitviewer.configuration import CONFIG as C
from aitviewer.renderables.smpl import SMPLSequence
from aitviewer.renderables.meshes import Meshes
from aitviewer.viewer import Viewer

if __name__ == "__main__":
    # Load a random AMASS sequence to get some SMPL data.
    c = (149 / 255, 85 / 255, 149 / 255, 0.5)
    seq_amass = SMPLSequence.from_amass(
        npz_data_path=os.path.join(C.datasets.amass, "ACCAD/Female1Running_c3d/C2 - Run to stand_poses.npz"),
        fps_out=60.0,
        color=c,
        name="AMASS Running",
        show_joint_angles=True,
    )

    # Extract two meshes, let's pretend vertices1 is the ground-truth.
    vertices1 = seq_amass.mesh_seq.vertices[0]
    vertices2 = seq_amass.mesh_seq.vertices[30]

    vertices1 = vertices1 - np.mean(vertices1, axis=0, keepdims=True)
    vertices2 = vertices2 - np.mean(vertices2, axis=0, keepdims=True)

    # Swap z and y because for AMASS z is up, but in the viewer y is up. This is usually not required if you don't use AMASS data.
    vertices1[:, [1, 2]] = vertices1[:, [2, 1]]
    vertices2[:, [1, 2]] = vertices2[:, [2, 1]]
    faces = seq_amass.mesh_seq.faces
    faces[:, [1, 2]] = faces[:, [2, 1]]

    # Pick a color map of your liking https://matplotlib.org/stable/users/explain/colors/colormaps.html
    from matplotlib.pyplot import cm
    color_map = cm.get_cmap('viridis', 256)
    color_map_output = color_map(np.linalg.norm(vertices2 - vertices1, axis=-1))

    # Create mesh objects and display it in the viewer.
    mesh1 = Meshes(vertices=vertices1, faces=faces, name="mesh1")
    mesh2 = Meshes(vertices=vertices2, faces=faces, vertex_colors=color_map_output, name="mesh2", position=np.array([-1.0, 0.0, -1.0]))

    v = Viewer()
    v.scene.add(mesh1, mesh2)
    v.run()

Which then looks like this (may be the cool color map would produce more similar results to your screenshot):

[wwwpkol]

Thank you very much for your enthusiastic and detailed answer!