At MWC 2026, AI voice technology provider iFlytek unveiled its eye-catching new wearable product – the "iFlytek AI Glasses." These smart glasses, designed specifically for cross-language face-to-face communication, not only boast an incredibly low weight of only 40 grams, but also pioneered multimodal noise reduction technology that combines lip movement recognition. When the other person speaks, real-time translated subtitles are displayed directly on the lenses, and the translated content is played back through the built-in speaker, achieving a truly "what you see is what you get" natural communication experience.
With the popularization of generative AI technology, smart glasses are regarded by the industry as one of the most ideal carriers for the practical application of AI. Many brands have begun...Entering the smart glasses marketFurthermore, with the addition of applications such as instant translation and head-up display, iFlytek, which has been investing in the translation field for a long time, is clearly preparing to apply its technological expertise to such vehicles.
Hands-free simultaneous interpretation assistant: multimodal translation and a lightweight design of 40 grams.
iFlytek AI Glasses combine two major features: "multimodal simultaneous interpretation" and "extreme lightweight design," eliminating the need for wearers to frequently look down at their phones for translation software during international conferences, business negotiations, or overseas exhibitions.
With the dual support of voice and visual translation, the glasses can instantly convert the other person's language into subtitles and project them in front of the eyes, while simultaneously playing them back as audio. To ensure comfortable wear for extended periods, iFlytek has put a lot of effort into the industrial design, keeping the weight of the entire pair of glasses to 40 grams, almost the same as ordinary optical glasses.
Killer technology: A multimodal noise reduction solution combining lip movement recognition
However, the most common fatal flaw of voice translation glasses in practical applications is that in high-noise environments such as exhibitions or parties, the microphone easily picks up too much background noise, resulting in a significant decrease in recognition and translation accuracy.
To solve this problem, iFlytek pioneered a multimodal noise reduction solution called "lip movement recognition". The system works by first capturing and tracking the speaker's lip movements in real time through a camera on the glasses, and then combining the wearer's voice captured by a bone conduction microphone to fuse the "lip movement image" and "audio" information.
Through this multimodal perception method, the glasses can accurately "lock in" and filter out the voice of the target speaker in a noisy background of multiple people talking. The official claim is that this technology can improve the accuracy of speech recognition and translation by more than 50%.
Listing Information
Currently, online pre-orders for the new iFlytek AI glasses have officially started, and local Chinese e-commerce platforms will fully open pre-orders at 10:10 AM on March 4th.
