Rough draft of satellite / active fire - #4
Conversation
|
|
||
| Researchers have built and meticulously annotated domain-specific datasets from this vast array of satellite imagery, such as [this dataset](https://drive.google.com/drive/folders/1GIcAev09Ye4hXsSu0Cjo5a6BfL3DpSBm) for space-based active fire detection introduced in [this paper by Pereira et. al](https://arxiv.org/abs/2101.03409). | ||
|
|
||
| While these datasets are information-rich and freely available, they are notoriously difficult to collaborate on for machine learning applications due to a few key factors: |
There was a problem hiding this comment.
I'm open to this being too much for now and better saved for a later "why satellite data is hard" style post
| **TODO IMAGE OF SUCCESSFUL FILE EXISTENCE CHECKS IN HUB** | ||
|
|
||
| #### 3. Improving Geographic Awareness | ||
| ##### Why emphasize geography? |
There was a problem hiding this comment.
-
Does this make sense? It's a very EO-specific concept but obviously related to ML as a whole...
-
Is it too much? I needed something to ground the geographical additions, and it's super pertinent to this dset, but it's also a pet fav topic of mine so don't want to ramble
There was a problem hiding this comment.
This feels like too much, could be a whole post on it's own. I think this post should focus on migrating from GitHub / Google Drive / Zip Files to Oxen, and why that is so much better.
|
|
||
| While the authors processed the full 1.6TB of imagery captured by LANDSAT-8 in August 2020 (targeting peak fire season), we're primarily interested in the excellent subset of data for which they manually created segmentation masks denoting active burn. | ||
|
|
||
|  |
There was a problem hiding this comment.
Can we put these images side by side? Feels like a lot of scrolling.
Also look at you being so nice to future generative AI data crawls with the alt text 😉
| **TODO IMAGE OF SUCCESSFUL FILE EXISTENCE CHECKS IN HUB** | ||
|
|
||
| #### 3. Improving Geographic Awareness | ||
| ##### Why emphasize geography? |
There was a problem hiding this comment.
This feels like too much, could be a whole post on it's own. I think this post should focus on migrating from GitHub / Google Drive / Zip Files to Oxen, and why that is so much better.
2 participants
@gschoeni still some refinement to do on my end re: wording and flow — want to get your thoughts on overall structure and story to make sure it makes sense and I'm not wandering out into the wilderness.
Working on cleaning up and documenting the analytic code as well.