TGArchive
·2 хв читання · 299 слів·👁 10.8K11

ELVIS V: The AI-powered Neuro Implant Restoring Vision

While the world buzzes about Elon Musk's Neuralink, the Sensor-Tech company is pioneering the world's first neuro implant, ELVIS V, capable of restoring sight. We caught up with company co-founder Denis Kuleshov to delve into the workings of this groundbreaking electronic vision technology.

👀 How ELVIS V Works

The patient wears a headband equipped with a camera that captures visual data. This information is processed, and the device transmits signals to the visual cortex in a language the brain can interpret. The electronics are implanted beneath the skin and attached to the skull. At the same time, an electrode-laden plate adheres to the visual cortex.

🤖 Understanding Perception with ELVIS

Images captured by the camera are sent to an information processing unit worn on the patient's belt. An AI algorithm processes these images, focusing on the contours of objects to relay only the most pertinent information to the brain. The data is then transmitted to the neuroimplant, enabling the individual to perceive the shapes of objects.

🏠 Object Identification System

The system's software, powered by AI, specializes in object classification. The camera captures multiple frames per second, and the offline neural network efficiently recognizes objects such as people, cars, and furniture.

🙈 Monkey Trials

In preliminary tests, monkeys are trained to recognize basic shapes using natural vision. Following the implantation of the neuro device, they perform tasks relying on electronic vision while wearing a special mask. The percentage of correct responses measures the success of these trials.

🤖 The Future of AI

By 2040, we expect to see a class of hybrid systems featuring a bionic component. This artificially grown intelligence will help achieve a breakthrough in AI. Additionally, we anticipate advancements in autonomous transport and intelligent assistants.

#interview | @hiaimediaen

Відкрити в Telegram
Повернутись до каналу