I have a set of custom images of a scene taken from different views. I would like to generate novel views based on pairs of reference image.
From my understanding, the current code is tied to specific datasets (e.g., co3d, re10k, etc.).
I'd love to get some guidance on how to adapt the code to work with my custom multi-view images.
Questions:
Generalization:
Assuming my dataset is very similar to re10k, is the checkpoint generalized enough?
Customization of Multi-View Images:
What specific customization do I need to make to my multi-view images in order to use them with your code?
Are there any specific formats or structures that my data should follow?
Scripts to Use:
Which scripts should I use to generate novel views from my reference images?
Are there any modifications required in the scripts to accommodate custom datasets?
Thank you for the amazing work on this project.
I have a set of custom images of a scene taken from different views. I would like to generate novel views based on pairs of reference image.
From my understanding, the current code is tied to specific datasets (e.g., co3d, re10k, etc.). I'd love to get some guidance on how to adapt the code to work with my custom multi-view images.
Questions:
Generalization: Assuming my dataset is very similar to re10k, is the checkpoint generalized enough?
Customization of Multi-View Images: What specific customization do I need to make to my multi-view images in order to use them with your code? Are there any specific formats or structures that my data should follow?
Scripts to Use: Which scripts should I use to generate novel views from my reference images? Are there any modifications required in the scripts to accommodate custom datasets?
Thank you again for this awesome work!