cellcanvas / cellcanvas/cellcanvas-prototype
feature: suggesting regions to label that are "hard to predict"
Open
- Dominant language
- Python
- Stars
- 12
- Forks
- 5
- PR merge metrics
- No merged PRs in 30d
Description
we can look into suggesting regions that are hard to predict by looking at where the prediction differs from the labels. We can then find similar pixels based on embeddings and suggest the user paint those regions.
Contributor guide
No contributing guide indexed for this repository
Research direction
Start by tracing where predictions, labels, and pixel embeddings are produced and compared. Done means the tool identifies regions where predictions differ from labels, finds similar pixels using embeddings, and suggests regions for the user to paint.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 25/100