BiAPoL / blog

Most of us study and work at the Bio-image Analysis Technology Development group at the DFG Cluster of Excellence “Physics of Life” at the TU Dresden. We blog about image data science, knowledge exchange and research data management in the life sciences.
https://biapol.github.io/blog/
Creative Commons Attribution 4.0 International
13 stars 8 forks source link

Improve yolo blog post #86

Closed jo-mueller closed 3 months ago

jo-mueller commented 3 months ago

@stefanhahmann pointed out that the upload of the processed image to the omero server leaves unclear where the target dataset id comes from. Maybe a short note on where this id comes from would be helpful.

It should also be noted that it's probably a bad idea to save the predictions back into the same dataset as it then cannot be distinguished which annotations are predictions and which are annotations.

There is also a wrongly formated note at the top.