What it is
A hybrid stack where lightweight models run on the glasses or phone and heavier models run in the cloud.
How it works
Wake-word detection and sensor fusion happen locally; images and audio snippets are sent for larger-model inference when needed.
Why it matters
Visual AI is the feature that turned smart glasses from an accessory into a category people actually buy.
Advantages
- Hands-free answers
- Live translation and captions
- Improves through software updates
Current limitations
- Privacy exposure
- Latency
- Battery cost of always-listening features
Commercial readiness
Shipping broadly in consumer AI glasses.
Development timeline
2024
Multimodal visual AI ships to consumer camera glasses