search

hbb1 / 2d-gaussian-splatting

[SIGGRAPH'24] 2D Gaussian Splatting for Geometrically Accurate Radiance Fields
https://surfsplatting.github.io
Other
2.13k stars 146 forks source link

Many surfels are like spikes #56

Open CanCanZeng opened 6 months ago

CanCanZeng commented 6 months ago

I used the code in issue https://github.com/hbb1/2d-gaussian-splatting/issues/55 to check the model generated by 2D-gaussian-splatting, and found that it has a common feature with the result of 3DGS, that is, many primitives are like spikes, rather than more intuitive ellipses. I don't know if the authors has discovered this phenomenon. I wonder if we can add a length-to-width ratio constraint to the scales of gaussian, which may lead to better reconstruction of the scene.

hbb1 commented 6 months ago

I think it will. I construct an extreme case.

without constraint image

with constraint (su==sv) (isotropic surfels) image

You can see that the distribution of isotropic surfels looks better. and there are less artifacts near the arm of the chair. Although such constraint may harm the PSNR, it has the potential to improve geometry. Or some soft constraints as you mentioned can be used for further trade-off. This is from my thought, not experimentally validated. Feel free to discuss.

CanCanZeng commented 6 months ago

Your experimental results are very encouraging, and I think we can delve deeper into this area

Zerui-Yu commented 2 months ago

You can see that the distribution of isotropic surfels looks better. and there are less artifacts near the arm of the chair. Although such constraint may harm the PSNR, it has the potential to improve geometry. Or some soft constraints as you mentioned can be used for further trade-off. This is from my thought, not experimentally validated. Feel free to discuss.

@hbb1 Greetings! Why do you think the PSNR will decrease by doing this? Could you please elaborate on your opinion?

renhaofan commented 2 months ago

You can see that the distribution of isotropic surfels looks better. and there are less artifacts near the arm of the chair. Although such constraint may harm the PSNR, it has the potential to improve geometry. Or some soft constraints as you mentioned can be used for further trade-off. This is from my thought, not experimentally validated. Feel free to discuss.

@hbb1 Greetings! Why do you think the PSNR will decrease by doing this? Could you please elaborate on your opinion?

Fig 10 of '3D Gaussian Splatting for Real-Time Radiance Field Rendering' and point-radiance may give you some insight.

Zerui-Yu commented 1 month ago

You can see that the distribution of isotropic surfels looks better. and there are less artifacts near the arm of the chair. Although such constraint may harm the PSNR, it has the potential to improve geometry. Or some soft constraints as you mentioned can be used for further trade-off. This is from my thought, not experimentally validated. Feel free to discuss.

@hbb1 Greetings! Why do you think the PSNR will decrease by doing this? Could you please elaborate on your opinion?

Fig 10 of '3D Gaussian Splatting for Real-Time Radiance Field Rendering' and point-radiance may give you some insight.

Thanks a lot! I finally figured it all out.