Question about evaluation code on spatial understanding and reasoning benchmark

[Tianbo-Pan]

Hi, thanks for sharing this excellent work and the codebase for G2VLM.

I noticed that the repository already provides the evaluation code for 3D reconstruction, but I could not find the evaluation code for the spatial understanding / spatial reasoning benchmarks.

Would it be possible to also open-source the evaluation code for the spatial reasoning benchmark?

[gordonhu608]

Thank you for your interests in our work. We follow evaluation benchmarks in their specific repos. At current stage, to evaluate our model, simply replace our model's inference code into the evaluation repo.
But we will try to organize the whole evaluation code as well :).