Vox-E: Text-guided Voxel Editing of 3D Objects

Supplementary Material

 

Please click on the sections below to view our interactive results, comparisons and ablations:

• Our Results
• Our Results on Extreme Edits
• Our Results on Real Scenes
• Comparisons
• Ablations
• Additional Ablations

 

---

Comparisons

In this section we compare to Text2Mesh and SketchShape.
We first show an experiment over uncolored meshes, as shown in Figure 6, followed by results on our dataset.
Note that Text2Mesh and SketchShape can only operate on uncolored meshes, and therefore they cannot faithfully reconstruct the input object (we visualize the input to each method on the left).

	Input	"A wood carving of a horse"	"A horse wearing a santa hat"	"A donkey"	"A carousel horse"
Text2Mesh
SketchShape
Ours

 

 

	Input	"A dog with a birthday hat"	"A claymation dog"	"A dog wearing a christmas sweater"	"A yarn doll of a dog"
Text2Mesh
SketchShape
Ours

 

 

	Input	"A cat with a birthday hat"	"A claymation cat"	"A cat wearing a christmas sweater"	"A yarn doll of a cat"
Text2Mesh
SketchShape
Ours

 

 

	Input	"A kangaroo with a birthday hat"	"A claymation kangaroo"	"A kangaroo wearing a christmas sweater"	"A yarn doll of a kangaroo"
Text2Mesh
SketchShape
Ours