Google Meet's model (image_segmenter)

[edamlmmv]

MediaPipe Solution (you are using)

image_segmenter

Programming language

JavaScript / WebGL2

Are you willing to contribute it

Yes

Describe the feature and the current behaviour/state

I am using the MultiSegment model as it is the most accurate (See: https://storage.googleapis.com/mediapipe-models/image_segmenter/selfie_multiclass_256x256/float32/latest/selfie_multiclass_256x256.tflite). However, it is not as accurate as Google Meet. I was wondering if there was any plan on releasing more accurate models?

Will this change the current API? How?

No response

Who will benefit with this feature?

Those who use image_segmenter from MediaPipe.

Please specify the use cases for this feature

Video conferencing / Artistry

Any Other info

No response

[kuaashish]

Hi @edamlmmv,

We apologize for the delay in our response. At present, we have not received an update from the team regarding the development of a more accurate model for image segmentation. However, I will note this as a feature request and share it to the team. Should there be any plans to address this matter, we will inform you through this same communication thread.

Thank you!!