OrCam Technologies introduced a device that isolates the voices of distinct speakers using their voice signatures. The device, OrCam Hear, is designed to amplify the targeted speakers, while effectively removing other voices and ambient noise for an improved audio experience.

The OrCam is powered by AI technology and proprietary voice enhancement deep learning models, emphasizing the company’s commitment to providing innovative AI-driven assistive technology for individuals with hearing challenges, according to the company.

OrCam Hear enables people to overcome the difficulty of understanding speech in noisy situations, known as the “cocktail party problem,” a well-recognized challenge for traditional hearing aids, the company says. The OrCam Hear helps people who are hard of hearing to discern and focus on specific voices in various social situations.

Further reading: Navigating Communication in a Noisy World

How the OrCam Works

The OrCam Hear’s EarBuds and a Mobile phone dongle are controlled by a dedicated app available for iPhone. After sampling the selected voices for a few seconds, using AI, the app creates a unique speaker profile that encapsulates the voice signature for each of the speakers. This allows users to selectively isolate specific voices even in noisy environments, according to the company. The OrCam Hear user can select the members of the conversation by turning each speaker “on”/”off” accordingly, with a single tap, enabling effective muting all other voices and ambient noise.

“The use of deep networks plus the latest network architectures large language models are harnessed to make a game changing experience for hearing aids and hearables in general,” says Professor Amnon Shashua, OrCam’s cofounder. “The problem of ‘speech in noise’ or commonly called the ‘cocktail party problem’ in academic literature has been a very difficult problem to tackle at a product level. Moving from an academic demonstration to a seamless product took years to perfect.”

OrCam Hear is currently in a technology preview phase and is expected to be shipped later in the year, the company says. More details about its release will be shared in the near future.

Photo: Dreamstime