Expose IrisToDepthCalculator in Python

[cancan101]

MediaPipe Solution (you are using)

No response

Programming language

Python

Are you willing to contribute it

Yes

Describe the feature and the current behaviour/state

IrisToDepthCalculator, which takes the iris landmarks and returns the depth of the iris from the camera is not currently exposed in Python. Iris landmarks themselves are now exposed in python and it would be great if these could then be converted to depths.

Will this change the current API? How?

No response

Who will benefit with this feature?

No response

Please specify the use cases for this feature

The ability to experiment in Python with iris depth estimation.

Any Other info

Related tickets:

• How use Mediapipe API for Python Iris Depth ? #2254
• Iris depth error in python  #2570

and would avoid developers from having to re-invent the wheel: https://github.com/Rassibassi/mediapipeDemos/blob/47c6330227c88ca02e32deb47b150aa8391306a4/custom/iris_lm_depth.py#L112-L123

[kuaashish]

Hi @cancan101,
Thank you raising this feature request. Could you please point out the use case about this feature that will be more helpful to understand the request better and we can bring up to notice of the team. Thank you!

[cancan101]

It can allow depth calculations without needing specialized hardware for the distance from the camera to a human face. For experimentation with these techniques it is helpful to have the algorithm exposed in Python rather than just in C++. As evidenced by the linked repository, given that the Iris depth logic has been (partially) re-implemented in Python, there is clearly a desire for this functionality in the community.

Linking to Google Blog Post the discussed the feature in general: https://ai.googleblog.com/2020/08/mediapipe-iris-real-time-iris-tracking.html.

[schmidt-sebastian]

Thank you for this feature request. We are a bit swamped at the moment but will add this to our backlog.