A vision for your app: On-device visual perception
Feed your app a photo and ask it questions, totally offline and without using any cloud services. Yes, your phone can do that now.
In this fully demo-led talk, we’ll see how to build powerful on-device vision capabilities using MediaPipe:
* Recognise hand gestures in a live video stream
* Feed images into Google’s new Gemma 3n model for complex visual question-answering
* Generate images
Create smarter, faster, and more private apps that can see and understand the world.
Session info:
Speaker: Tom Colvin
Android GDE, freelancer, co-founder of Apptaura at Apptaura
Date: 13 March 2026
Time: 09:55 - 10:40
Relevant tags:
Android
