Skip to content

Releases: aubricot/computer_vision_with_eol_images

0.2.0

24 Oct 22:13
@aubricot aubricot
v0.2.0
704421b
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.
Compare
Choose a tag to compare
0.2.0 Pre-release
Pre-release

This release of Computer Vision With EOL Images runs with TensorFlow 2.18, and YOLO v3 in darknet in Python 3.

Major Features and Improvements

  • Many functions are now pushed to external python modules to tidy up Colab notebooks.
  • Version control via requirements.txt files was established to enable full rollbacks and avoid issues caused by Colab and Tensorflow background updates.
  • Most important outputs are color-coded (Red: possible error, Yellow: warning, Green: success) to facilitate user interaction with code.
  • Demo scripts are now available under play_for_beginners.
  • Feature request template was added.

Bug Fixes
[Fix] Training using model_main_tf2.py no longer works in TF2.8+ bug (#9)
[Fix] saved model doesn't work with newest cloned version of TF Objdet API bug (#7)
[Fix] Objdet for image cropping with EOL saved model doesn't work after TF update (#9)

Assets 2
Loading

0.1.0

31 May 19:29
@aubricot aubricot
v0.1.0
b3db6e6
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Verified
Learn about vigilant mode.
Compare
Choose a tag to compare
0.1.0 Pre-release
Pre-release

This release of Computer Vision With EOL Images runs with TensorFlow 2.8 and YOLO v3 in darknet in Python 3.

Parameters can be modified through form fields.
Object detection tagging and cropping notebooks can be run as demos in Colab runtime without connecting to Drive.
All notebooks can be run independently. Entire repository does not need to be cloned. Needed files are downloaded within each notebook.

Assets 2
Loading