eyemoreX42 imaging chip released to fill the gap in the global AI vision chip

It is understood that the eyemore imaging chip is a 100% self-developed visual imaging chip of Eye Engine, which aims to "provide an AI vision with an imaging engine that clearly sees the world." The imaging engine is a dedicated imaging chip based on ultra-large-scale computing, self-developed imaging algorithms and over 500+ scene data. It mainly solves the core of high-quality imaging of complex light such as low light, backlight and reflection in AI vision. Pain points. The eyemoreX42 eliminates the interference of live light in a variety of complex lighting environments, and outputs stable and reliable high-quality visual images to the AI ​​vision algorithm. The imaging engine simulates the response of the eyeball to light, intelligently processes the contrast light with super-high latitude, and produces rich and delicate layering and color-sharp contrast, and outputs high-quality photo image.

In the past 4 years, Eye Engine has been committed to the development and iteration of eyemore imaging chips. Zhu Jizhi, CEO of Eye Engine Technology, found that in the process of machine vision processing, in addition to the cognitive analysis of the back end, it is responsible for handling various Complex light scenes, front-end perception systems that generate high-quality images are also essential, equivalent to the "eyes" of the AI ​​cognitive world. And if the domestic AI field wants to get a lot of development, this "eye" perception and imaging is a technical "gap." At present, eyemore will focus on four market application directions: visual imaging for autonomous driving; AI imaging for smartphones; high-end intelligent security based on face recognition; industrial visual imaging including industrial monitoring and medical.

The actual cooperation project of Eye Engine has covered the application of domestic face recognition security , automatic driving, medical, industrial monitoring, railway transportation and other major fields. Among them, the high-end security industry extended by AI face recognition is a very promising market. The imaging engine provided by eyemore can directly help visual AI companies to obtain high-quality image data, thereby improving the recognition rate; for the autopilot market, eyemore plans to jointly Automated driving providers in the field develop car-level imaging modules and chip products. In the mobile phone market, although mobile phone manufacturers have a longer Design-in cycle, more mobile phone manufacturers have begun to use AI camera capabilities as a mobile phone. With the new standard features, eyemore's imaging technology is expected to become an indispensable technical support for manufacturers.

When talking about the biggest advantage of AI imaging technology, Zhu Jizhi said: "The eyemore team has built a complete software and hardware solution for visual depth learning and cognition for computer vision providers, including AI imaging engine and the engine. Encapsulation to the hardware side of the imaging module, but we believe that the future is still a "software-defined hardware" era, so our imaging engine architecture is designed to be very easy for visual AI software developers to quickly learn to use."

In the existing application case of Eye Engine, eyemore partner developed an embedded AI camera for school classroom scenarios and based on face recognition technology to calculate student attendance in the classroom; the team also cooperates with design companies in the e-commerce field. Developed a depth camera equipped with structured light (used to collect three-dimensional information of the object) and an AI imaging module to ensure that the acquired three-dimensional image is the same color as the real object.

At the end of the conference, Eye Engine demonstrated the AI ​​imaging technology. At the scene of the conference, the image of the human eye is close to the darkness. Under the premise that the ordinary AI algorithm is unrecognizable, the equipment equipped with the eye engine technology engine can still restore the image clearly, and the algorithm also clearly recognizes the shooting environment. Everyone and information about the object.