We select one random scene per category, from a subset of CO3D, and show comparison against PixelNeRF, NerFormer, and ViewFormer.

We select one random scene per category and show results with varying number of input views.

We select one random scene per category and show comparison against NerFormer and ViewFormer.