You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I hope this message finds you well. I have been using the 3dgs library for training and aligning 3D geospatial models in the Earth-Centered, Earth-Fixed (ECEF) coordinate system. While experimenting with the library, I have encountered a significant performance drop in my models after aligning them in ECEF compared to their pre-alignment state.
Questions:
Parameter Tuning:
Is it possible that the performance drop can be mitigated by adjusting certain parameters in the training or alignment process?
If so, could you suggest which parameters to tweak and provide any guidance on how to do so effectively?
Order of Operations:
Is the approach of training the models first and then aligning them to ECEF correct, or should the alignment be done prior to training?
Alternatively, is there a recommended workflow that combines training and alignment in a way that preserves model performance?
Thank you very much for your time and assistance. I look forward to your response.
The text was updated successfully, but these errors were encountered:
I hope this message finds you well. I have been using the 3dgs library for training and aligning 3D geospatial models in the Earth-Centered, Earth-Fixed (ECEF) coordinate system. While experimenting with the library, I have encountered a significant performance drop in my models after aligning them in ECEF compared to their pre-alignment state.
Questions:
Parameter Tuning:
Is it possible that the performance drop can be mitigated by adjusting certain parameters in the training or alignment process?
If so, could you suggest which parameters to tweak and provide any guidance on how to do so effectively?
Order of Operations:
Is the approach of training the models first and then aligning them to ECEF correct, or should the alignment be done prior to training?
Alternatively, is there a recommended workflow that combines training and alignment in a way that preserves model performance?
Thank you very much for your time and assistance. I look forward to your response.
The text was updated successfully, but these errors were encountered: