Release Date
12 23, 2025
When AI evolves from understanding speech to understanding what users see, human-computer interaction enters a new multimodal stage.
On December 23, Union Image and Guangfan Technology jointly launched the world's first vision-aware active AI earphones, opening a new chapter for scene-aware interaction and wearable AI hardware.

Dual-Camera Vision System for AI Wearables
As the visual system partner, Union Image provided the left and right ear dual-camera vision system for Guangfan Lightwear AI wearable devices, integrating stable, mass-production-ready visual capability into an extremely compact form factor.
The module uses a 2MP sensor with 1600 x 1200 effective resolution, 1.75um pixel size, and a compact 8 x 10.31 x 4.01 mm structure, balancing power consumption, image quality, and mechanical space.

From Listening to Seeing: Multimodal AI Interaction
The camera is not designed for simple photography. It is a key entry point for environmental understanding and contextual perception. By combining vision, voice, and other multimodal information, AI earphones can better understand user intent, identify targets, plan tasks, and call services based on real-time scenes.
The solution integrates ISP and scanning capability, supporting high-quality still images, low-power dynamic capture, and fast response for high-frequency interaction scenarios such as payment and information access.

Ultra-Miniature Camera Modules for Smart Wearables
Smart wearable vision solutions must solve space, thickness, image stability, power consumption, and mass-production challenges. Union Image continues to develop ultra-miniature camera module solutions for wearable devices and AI glasses.
A flagship option uses the Sony IMX681 sensor with 12MP resolution, supporting first-person recording, AI recognition, and multimodal interaction with sufficient image detail. Ultra-wide DFOV options from 104 to 114 degrees support environmental perception, content creation, and AI understanding.

For smart wearables, the true challenge is not whether a camera exists, but whether visual capability can be delivered reliably within extreme space constraints. This is where Union Image's full-chain capability in optical design, hardware engineering, algorithm integration, system delivery, and intelligent manufacturing becomes essential.
Union Image will continue working with partners to bring AI vision into more intelligent terminal forms, connecting AI with real-world perception and accelerating multimodal experiences for everyday life. Vision for All, AI for Life.



