Implement Explainable AI for 'Object detection/classification' over existing and new models.

[hemantkshirsagar8]

Task.

• Implement Explainable AI for 'Object detection/classification' over existing and new models.
• For object detection, use tensorflow, scikit-learn and fatst-ai.
• Can take any sample use case data from Scikit-learn examples.
• Features need to be self-explanatory during model predictions.
• Bind model in ML pipeline using MLFlow.

[hemantkshirsagar8]

Hows the progress, any difficulties?

[hemantkshirsagar8]

On the befalf of student;

• Tried one example from scikit learn for coin classification.
• Will Integrate with SHAP.
• Come up with results and understanding.

[Eshwar7799]

https://www.youtube.com/watch?v=C80SQe16Rao

[Eshwar7799]

https://scikit-image.org/docs/dev/auto_examples/

[hemantkshirsagar8]

• We concluded that existing code for coin detection does not use ML techniques.
• Please google or explore further to get any Sci-kit based sample code for ML model training, building, and training. (Avoid tensorflow code for now)

[hemantkshirsagar8]

Any progress on this?