VeriSilicon (688521.SH) has announced the launch of its latest AI-driven image processing intellectual properties (IPs), the AI-NR and AI-SR series, designed to deliver superior noise reduction and super resolution capabilities. These new solutions, including the AINR1000 and AINR2000 for intelligent noise reduction and the AISR1000 and AISR2000 for advanced super resolution, are optimized for performance, power efficiency, and area (PPA). They offer a cost-effective and highly scalable approach to image enhancement, making them ideal for a wide array of applications, including automotive imaging, surveillance, cloud gaming, and consumer electronics.
The AI-NR series utilizes VeriSilicon’s proprietary AI pixel processing algorithms to enhance image clarity by effectively reducing noise while preserving intricate details and color accuracy. Designed to function in challenging lighting conditions, the AINR1000 provides a resolution of 2560×1440 at 30 frames per second (fps), making it suitable for cost-sensitive applications such as entry-level Internet Protocol Cameras (IPCs). Meanwhile, the AINR2000 supports a higher resolution of 5600×4208 at 60fps, delivering superior performance for advanced surveillance systems, sports cameras, and other high-end imaging solutions.
VeriSilicon’s AI-SR series brings AI-powered super resolution to the forefront, enabling high-quality upscaling of Standard Definition (SD), High Definition (HD), and Full High Definition (FHD) content to 4K. Leveraging advanced operator-layer combinations, these IPs refine details, eliminate jagged edges, and produce sharper, more natural images. The AISR1000 supports 4K upscaling at 60fps, while the AISR2000 features a dual-engine architecture, achieving 4K at up to 144fps, ensuring both high performance and efficient power consumption. The AI-SR series offers flexible scaling ratios of up to five times, making it an optimal solution for real-time high-resolution applications in cloud gaming, virtual cloud desktops, and AI-enhanced video streaming.
VeriSilicon’s FLEXA interface communication technology enables seamless integration of these IPs with the company’s existing Image Signal Processing (ISP), Video Processing Unit (VPU), and Display Processing IPs. Additionally, they can be incorporated into third-party imaging pipelines, offering system designers greater flexibility in deploying AI-powered image enhancement across a variety of platforms.
Weijin Dai, Chief Strategy Officer, Executive Vice President, and General Manager of VeriSilicon’s IP Division, emphasized the company’s commitment to advancing AI-driven imaging solutions. “Our AI-NR and AI-SR series IPs are engineered to significantly enhance image quality while maintaining low power consumption and high silicon efficiency. These technologies have already been deployed in multiple customer applications, including battery-powered cameras, mobile application processors, and data centers for AI-enhanced video transcoding. As we move forward, VeriSilicon remains dedicated to expanding our portfolio in both image reconstruction and image enhancement technologies, continuously introducing new capabilities to improve image quality and system efficiency.”
With the launch of these AI-NR and AI-SR IPs, VeriSilicon continues to drive innovation in AI-based image processing, providing industry-leading solutions that enhance visual clarity and performance across a broad spectrum of digital imaging applications.