KuangshiAi
merge chatvis bench and main
77dc289
|
raw
history blame
706 Bytes

Side-by-side comparison of an isovolume featuring clipping, labeling, and color mapping

  • Sets up a side-by-side layout with two render views to compare a prediction (neg10_prediction.vtr) vs ground-truth (neg10_ground_truth.vtr) dataset.
  • For each, converts cell data (Intensity, Phase) to point data, extracts an isovolume on Intensity in [0.2, 1.0], then applies a clipping plane at origin (32,32,32), and colors the result by Phase using a Viridis colormap with visible scalar bars.
  • Adds text overlays (“NN Prediction” and “Ground Truth”), aligns both cameras to −Z with a 45° yaw for consistent views, and saves a combined screenshot of the layout to comparison-screenshot.png.