AXERA Shines at the 2022 World Artificial Intelligence Conference, Layout in Three Major Industries

Release time：2022-09-01

To demonstrate the superior imaging capabilities of AI-ISP beyond traditional norms, AXERA has established a darkroom laboratory within the exhibition area of this AI conference. Through real-time comparative testing, Aixin Zhimou® AI-ISP's imaging performance under extremely low-light conditions is showcased. The laboratory highlights the exceptional features of Aixin Zhimou® such as reducing motion blur, color restoration, and low-light noise reduction.

AXERA is also the pioneer in achieving commercial implementation and large-scale production of AI-ISP technology on chips. Currently, they have successfully mass-produced two generations of five edge AI chips. At this exhibition, AXERA's complete series of chip products garnered significant attention due to their impressive performance, capturing the audience's interest.

Multi-Domain Product Deployment, Focusing on New Intelligent Consumer Experiences

With the continuous innovation of technology, the application scenarios for AI chips have been increasing. In terms of commercial deployment, AXERA's core products and application solutions have been widely applied in various fields such as smart cities, intelligent consumer products, intelligent transportation, and wearable technology. They have been consistently expanding their applications in both B2B and B2C domains. At this exhibition, AXERA focused on showcasing their successful implementations in three major areas: smart cities, intelligent consumer products, and intelligent transportation, providing a comprehensive view of their broad AI chip industry ecosystem.

In the domain of intelligent consumer products, AXERA's specialized imaging enhancement chip AX170A successfully passed the acceptance test of a prominent smartphone client earlier this year, marking AXERA's successful entry into the consumer electronics market. During the exhibition, they demonstrated the 'Night Scene Enhancement' photo effect based on the AX170A chip in smartphones, and they also displayed physical examples of underwater robots and expandable cameras. The underwater robot is equipped with AXERA's AX620A chip, capable of achieving underwater AI low-light full-color capabilities and AI-based clarity enhancement, ensuring clear underwater imaging. The expandable camera, based on the AX620U module with its powerful computing capabilities, can be widely used in AIoT scenarios.

In the context of smart city development, AXERA leverages their AI visual core technology to process video streams in a structured manner, functioning as the 'eyes' of smart cities. In this exhibition, AXERA presented the industry's first real-world demonstration of AI-ISP dual-camera fusion products, which can achieve high-color fidelity and reproduce natural details even in extremely low-light conditions of 0.01 lux. They also showcased an embedded edge computing box with impressive capabilities of 28.8 teraflops of computing power and the ability to process 16 video streams simultaneously, enabling intelligent city management for enhanced urban safety.

In the field of intelligent transportation, the AI-powered black light BSD intelligent camera, equipped with the AX620A chip, can operate around the clock. It particularly excels in intelligently detecting obstacles in vehicle blind spots, such as pedestrians, cyclists, road edges, and barriers, during low-light nighttime scenarios, providing alerts and warnings. The 4K dashcam, also equipped with the AX620A chip, offers features like 4K black light night vision, distance warning for following vehicles, traffic light reminders, and pedestrian recognition alerts. AXERA's technical spokesperson stated that they are committed to utilizing the product's capabilities of high image quality, full intelligence, complete perception, and all-weather functionality to ensure travel safety, enhance traffic efficiency, and effectively contribute to the upgrade of intelligent transportation.

Building the AX-Pi Open Source Ecosystem to Make Edge AI Everywhere

Based on the self-developed AX620A chip, AXERA, in collaboration with multiple partners, has initiated an ecosystem co-creation plan. On-site, AXERA showcased their open-source ecosystem development board called 'AiXin Pi' (AX-Pi), which supports various scenarios and devices such as color recognition, vehicle identification, and facial detection, making edge AI ubiquitous.

Around the AiXin Pi hardware development board, AXERA also provides system software and documentation to support AX-Pi's usage, thus fostering an open-source ecosystem. They have launched bounty tasks within the community, encouraging developers to submit interesting samples to the AX-Samples open-source project and address any issues in code or documentation. This approach enhances the collaborative efficiency and expands the algorithmic ecosystem of domestically developed chips. At the event, the intelligent tracking camera developed based on AiXin Pi garnered significant attention, with many attendees trying out the skeleton recognition technology.

The deep integration of AI and the visual industry is greatly propelling the upgrade of intelligent terminals. This integration plays a crucial role in optimizing consumer experiences, improving social efficiency, and visual AI has become one of the most prominent research areas in artificial intelligence. Looking ahead to the rapidly developing AI industry, AXERA will continue to focus on the development of 'perception + computation,' accelerating the mass production and application of AI visual perception chips, and paving the way for the future by unleashing limitless possibilities in the visual realm.