Gemma3n multimodal(image,audio) support for Android and web

[bil-ash]

Have I written custom code (as opposed to using a stock example script provided in MediaPipe)

No

OS Platform and Distribution

Android, Web

MediaPipe Tasks SDK version

No response

Task name (e.g. Image classification, Gesture recognition etc.)

ASR, image recognition

Programming Language and version (e.g. C++, Python, Java)

Java, js

Describe the actual behavior

Mediapipe is able to accept only text and image input on Android and only text input on web.

Describe the expected behaviour

Mediapipe is able to accept text,image and audio input on both Android and web

Standalone code/steps you may have used to try to get what you need

Followed the official docs

Other info / Complete Logs

[kuaashish]

Hi @bil-ash,

Thank you for raising this request. We are forwarding it to our team for review. Hopefully, support for image and audio in the new Gemma 3n through MediaPipe will be available soon.

[tp3344]

@kuaashish
Did team Mediapipe forget to build tasks-genai for android with version 0.10.26? It contains functionality to run llm inference with audio as far as I can see.