-
Notifications
You must be signed in to change notification settings - Fork 10
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Inference for downstream tasks #18
Comments
Hi, thanks for the interest in our work! Since we synthesize RGB images of whole objects (i.e. perform amodal completion and segmentation), our approach makes it straightforward to equip various computer vision methods with the ability to handle occlusions, beyond amodal segmentation. After performing amodal completion, for recognition, we use CLIP as the base open-vocabulary classifier. For novel view synthesis and 3D reconstruction, we use SyncDreamer. Since these codebases are very clean & well-documented, we don't plan to provide the intermediary code here. Moreover, please note that our approach is not specific to any particular recognition, or NVS/3D reconstruction model. Instead, it serves as a drop-in module to enable them to handle occlusions. That said, please feel free to contact me at ege.ozguroglu@columbia.edu for specific scripts. |
Thanks for the response! |
Hello,
Do you plan to provide inferencing for the downstream applications mentioned, such as 3D reconstruction from occluded RGB image?
The text was updated successfully, but these errors were encountered: