Demo visual results differ significantly from paper figures

[chiller-chen]

Hi, thanks for the excellent work and for releasing the demo.

I tried running several LL3M demo scripts following the repository instructions. However, the generated visual results differ noticeably from the figures shown in the paper. I would like to confirm whether this discrepancy is expected or if I may have misconfigured something in the workflow.

Below are several examples:

[trospix]

Yes I had the same impression, but the overall resuls is better than blender-mcp

Have you tried others like 3D-Agent ? It also focus on clean wireframes different from like Meshy or Rodin.

[hcaumo]

Yes I had the same impression, but the overall resuls is better than blender-mcp

Have you tried others like 3D-Agent ? It also focus on clean wireframes different from like Meshy or Rodin.

3D-Agent is the only that really worked for me