White canes are fine for helping blind users make their way down streets and hallways, but they certainly can't tell those users what objects, signs or scenes are in front of them. That's exactly what the Seekr is designed to do, however, using a heapin' helping of AI.
The Seekr device, which is being commercialized by Hong Kong startup Vidi Labs, was a recent recipient of a CES 2025 Innovation Award.
Vidi Labs director Turzo Bose was inspired to create the gadget in 2020 during the pandemic. He realized that due to social distancing precautions at the time, his visually impaired grandfather couldn't get together with sighted people who would describe things to him while out and about. What his grandpa needed was a device that would do the job.
The result is the matchbox-sized Seekr, which can be clipped onto the user's clothing or held in one hand.
It's equipped with a depth-sensing camera, microprocessor, battery and Bluetooth module, among other electronics. The device is paired with a paid iOS/Android app on the user's smartphone, which in turn relays synthetic voice cues via a third-party Bluetooth earpiece.
Users can switch between four AI-based operating modes, all of which are activated by pressing a button on the Seekr to take a photo.
Scene mode, appropriately enough, describes the basic scene in front of the user. The voice cue might say "Two people sitting on a bench at your 12 o'clock," for instance.
Depth mode, on the other hand, guides the user's hand toward an item that they're trying to grasp. Supermarket mode identifies the type of product photographed (bread, apples, etc), and Text mode reads out text on signs, packaging, menus, etc.
The device can be ordered via the Seekr website, and is priced at 6,000 Hong Kong dollars (approximately US$770) – that price includes a one-year subscription to the app.
Source: Seekr
