own data do not give good 3D reconstruction
- Dominant language
- Jupyter Notebook
- Stars
- 2.8k
- Forks
- 458
- PR merge metrics
- No merged PRs in 30d
Description
Hello there, I tried your code to test on my own data. You can find my training image example as follows,
I collected 50 images to cover 360-degree view. I used your code for reconstruction, the results are like this,
There are some holes and blur on the background.
As you said, some background can help for 3D rendering, but apparantly, NeRF cannot handle objects with complicated background, right? Can you share your opinion to us? Thank you!
Contributor guide
No contributing guide indexed for this repository
Research direction
No source file, test, or entry point is named. Start by reviewing the supplied training-image and reconstruction examples, then reproduce the result with the reported 50-image, 360-degree capture; done means establishing whether the holes and background blur are a reproducible defect or expected behavior.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- pytorch
- Domain
- computer-vision, machine-learning
- Issue type
- Bug
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100