facebookresearch / facebookresearch/segment-anything

Aim Integration

Open
#134 0 comments 1 reaction 0 assignees View on GitHub

Nobody has claimed this yet.

discussion enhancement
Dominant language
Jupyter Notebook
Stars
54.9k
Forks
6.4k
PR merge metrics
No merged PRs in 30d

Description

🚀 Feature

Integration with Aim.
It would be great to have an easy way of tracking metadata on Aim.

Motivation

To enhance experiment reproducibility and provide a unified, aesthetically pleasing user interface for exploring samples and their corresponding segmentations, it is desirable to furnish users with a solid tool. Aim facilitates gaining deeper insights, thus enhancing the quality of experimentation.

Pitch
  • Retrieve metadata of the model, dataset, and annotation from a given segment predictor instance.
    • Store fields such as image_info , annotation, model, etc.
  • Store the metadata in Aim run
  • Display stored metadata in run page on Aim UI.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

The issue names no files, tests, or entry points. Start by locating the segment predictor and the model, dataset, and annotation metadata it exposes, then review how an Aim run stores and displays metadata. Done means those metadata fields can be stored in an Aim run and viewed on the Aim UI run page.

Written by the indexing model from the issue text.

Assessment

Domain
machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.