Hi, thanks for your novel framework of neural 3d reconstruction.
I'm interested in the generalizability of NeuralRecon, but when I tried to conduct NeuralRecon on NYUDepthv2, poor prediction meshes are achieved but the generated gt meshes are correct. Are there any parameters needed to adjust on other datasets? I can upload several scenes of the NYU dataset if needed.
Hi, thanks for your novel framework of neural 3d reconstruction.
I'm interested in the generalizability of NeuralRecon, but when I tried to conduct NeuralRecon on NYUDepthv2, poor prediction meshes are achieved but the generated gt meshes are correct. Are there any parameters needed to adjust on other datasets? I can upload several scenes of the NYU dataset if needed.