About visual prompt object queries
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 1.4k
- Forks
- 55
- PR merge metrics
- No merged PRs in 30d
Description
For the interactive mode, following SAM, which employs prompt encoder to encode various visual prompts as location queries.
This has the limitation that it relies mainly on spatial information to guide the segmentation process and cannot query and segment the same visual target in another image.
Is it possible to similarly encode visual prompt as semantic queries like text to handle the segmentation of the same visual target across images?
Contributor guide
No contributing guide indexed for this repository
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 interactive mode, SAM, and prompt encoders but no files, tests, or entry points. Start by locating the interactive segmentation flow and its visual-prompt encoding path, then compare how prompts are represented across images. Done would require a defined semantic-query design and evidence that the same target can be segmented across images.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- computer-vision, machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100