Add streamlit page for object detection, segmentation, and pose extraction

DecodingRaphael / unraphael

Python toolkit for unraveling image similarity with a focus on artistic practice

https://unraphael.readthedocs.io

Apache License 2.0

2 stars 1 forks source link

Add streamlit page for object detection, segmentation, and pose extraction #46

Closed ThijsVroegh closed 3 months ago

ThijsVroegh commented 3 months ago

PR for additional page to the streamlit app, covering the detection, segmentation and pose extraction of individual objects in an image. Allows for saving individual objects. This could be used in the analysis pipeline, in case the overall outline of figures is not being used, but instead focus lies on the structural similarity of an individual figure.

TODO:

[x] Adapt the code so that not all the output dissapears when an image is downloaded.

ThijsVroegh commented 3 months ago

@stefsmeets I'm happy with the current functionality of this page, except for the one TODO thing listed in the PR.

stefsmeets commented 3 months ago

Cool, I will have a look at it tomorrow or Wednesday.

stefsmeets commented 3 months ago

Made a first pass through the code, let me know if you are happy with the changes, then I will finalize it so that it can be merged.

ThijsVroegh commented 3 months ago

Nice @stefsmeets , looks very good to me! Thanks!