Hi there @aras-p ,
Your effort in bringing the original pre-trained 3D Gaussian Splatting (3DGS) model into Unity has been very impressive and extremely helpful to the community.
For my current research project, I would like to extend your implementation to support omnidirectional (360°) 3DGS models that are trained purely on omnidirectional imagery (instead of perspective images). However, as I am still quite new to the codebase, I’m not sure where would be the best place to start making such modifications.
I would greatly appreciate it if you could kindly share some guidance or suggestions — for example, which parts/files/classes you think are most relevant to look at first when trying to adapt the pipeline for omnidirectional input.
Thank you very much in advance for your time and for any advice you’re willing to share.
Best regards,
Gia Hieu
Hi there @aras-p ,
Your effort in bringing the original pre-trained 3D Gaussian Splatting (3DGS) model into Unity has been very impressive and extremely helpful to the community.
For my current research project, I would like to extend your implementation to support omnidirectional (360°) 3DGS models that are trained purely on omnidirectional imagery (instead of perspective images). However, as I am still quite new to the codebase, I’m not sure where would be the best place to start making such modifications.
I would greatly appreciate it if you could kindly share some guidance or suggestions — for example, which parts/files/classes you think are most relevant to look at first when trying to adapt the pipeline for omnidirectional input.
Thank you very much in advance for your time and for any advice you’re willing to share.
Best regards,
Gia Hieu