tensorflow / tensorflow/graphics
CVXNet Training with custom data for 2D case
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 2.8k
- Forks
- 374
- PR merge metrics
- No merged PRs in 30d
Description
Hi @Awcrr
I'm trying to reimplement the cvxnet model in pytorch and train for a custom dataset of images. I intend to do single view reconstruction but for the 2D case i.e output 2D polygons/shapefiles instead of 3D meshes. I have a custom dataset of images and corresponding gt raster seg masks. Can you please tell me how to preprocess this data for training with cvxnet?
In particular, I'm confused about how to generate the sample points and the corresponding point labels for this dataset.
If I have to guess, I sample random coordinates from the image array (x, y) and get their corresponding label from the raster gt seg masks. Is that correct?
Any help will be appreciated. Looking forward to your reply.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
The issue names CVXNet, PyTorch, custom image data, and raster ground-truth masks, but provides no files, tests, or entry points. First locate the existing CVXNet sample-point and point-label preprocessing path, then compare it with the proposed image-coordinate and mask-label workflow. Done means clear guidance for preparing 2D training data.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- computer-vision, machine-learning
- Issue type
- Documentation
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100