tensorflow / tensorflow/tensorboard

Feature request: Support image metadata in embedding visualizer

Open
#377 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

type:feature
Dominant language
TypeScript
Stars
7.2k
Forks
1.7k
Avg merge
4d 22h
Merged PRs (30d)
1

Description

I'd like to use TensorBoard to visualize an embedding for training data. I have ~100K images, that are 1000x1000. The images are too large for the sprite thumbnails -- as per the docs, it appears that the sprite loader supports a total resolution of 8192 x 8192.

Is there a way I could attach my images to the metadata card?

While the existing metadata.tsv parser only supports text data, it seems like I could achieve this functionality by editing the HTML here. It would be nice to have a more elegant solution -- perhaps by linking metadata to existing TensorBoard image summary visualizations.

Any suggestions / comments are appreciated, thanks.

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

Start with tensorboard/plugins/projector/vz_projector/vz-projector-metadata-card.html near the referenced line and read the embedding visualizer documentation linked in the issue. Clarify how image metadata should connect to existing TensorBoard image summaries and define completion as displaying the requested training images in the metadata card without relying on the sprite resolution limit.

Written by the indexing model from the issue text.

Assessment

Tech stack
html, typescript
Domain
data-visualization, frontend
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.