Issue 1

Open
#1 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Assessment

Difficulty
5/5
Estimated time
Over a week
Newbie friendliness
20/100
Issue type
Feature
Clarity
Needs clarification
Activity status
Stale
Tech stack
aws, jupyter-notebook, machine-learning
Domain
cloud, data-engineering, machine-learning

Research direction

Start by reviewing the linked labs issue and the listed Phase 1 milestones. The payload identifies no repository files, tests, or entry points. Completion is not defined beyond planning pipelines, generating embeddings, storing them as Zarr on Source Cooperative, and providing downstream examples.

Written by the indexing model from the issue text.

Description

issue-1

Goals

  • Help increase ease of access to various embeddings in cloud native formats

  • Provide an example of what cloud native easily accessible, interoperable embeddings from Geospatial Foundation Models could look like by storing embeddings derived from 3 different models on Source Cooperative for 8 years (2017-2024) for all a sample geographic area (ie all of Kenya)

  • Make it easier to use embeddings in downstream modeling classes

Partner and Stakeholders

Taylor Geospatial Engine, Fields of the World, Open Source Developers: Issac Corley, Jeff Albrecht, Source Cooperative

Team

Dev Seed Gaia + other labs contributors

Milestones

Phase 1 November - December 2025

  • Simple + scalable engineering AWS pipeline for S2 and S1 download needed for embedding generation

  • simple + scalable engineering AWS pipeline for embeddings generations for models we want to run inference with

  • Experiments to guide initial decisions around: zarr file chunking, projections

  • Generate (when necessary) embeddings, store as Zarr - determine chunking scheme host on source cooperative

Phase 2 (Application Based) December/Jan 2025

  • tools + examples to facilitate comparison for downstream use cases (ie field boundary delineation) with the embeddings that are stored on source cooperative

Future Work:

  • Enable easier visualization + visual comparison of different model's embeddings

  • Connect Embeddings to datasets/stac collections that they are derived from

  • Going beyond country level scale - do our initial decisions hold out at areas larger than a country

Project Organization

  • slack channel tbd
  • project board - coming soon 🚧

Documents

https://github.com/developmentseed/labs/issues/458

cc @zacdezgeo

Dominant language
Jupyter Notebook
Stars
32
Forks
0
No merged PRs in 30d

Contributor guide

No contributing guide indexed for this repository

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.

More from developmentseed/pixelverse

All issues in developmentseed/pixelverse

Similar issues

More Cloud issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.