CEVA and ASUSTEK Technologies present voice experience in second generation smart camera and video conferencing system

CEVA and ASSETS Collaborate to Enable Exceptional Voice Experience on Cupola360 SoC for Second Generation Smart Camera and Video Conferencing Systems

CEVA-BX1 DSP Enables Audio/Voice Capabilities of ASUN AST1230 Smart Camera SoC; ASUN Customers Can Use CEVA ClearVox Voice Front-end Software to Implement the Most Challenging Multi-Microphone Conferencing Use Cases

CEVA, the world's leading licensor of wireless connectivity and smart sensing technologies and integrated IP solutions and ASPEED Technology announced that ASPEED Technology has licensed the CEVA-BX1 audio/voice DSP for use in the second-generation Cupola360 SoC AST1230 for smart cameras and video conferencing systems. Through CEVA's ClearVox multi-microphone noise reduction and echo cancellation audio front-end (AFE) software, the two companies built solutions to the most challenging conferencing system use cases. This software package is fully optimized for the CEVA-BX1 DSP, greatly improving the clarity of any voice conferencing system and allowing the addition of voice assistants and hands-free control capabilities.

"The second generation Cupola360 SoC is our first to integrate real-time multi-image stitching video and powerful audio processing capabilities, making it perfect for video conferencing applications," commented Lin Hongming, Chairman and CEO of ASINNOVA. "Through our full collaboration with CEVA, our customers can enjoy the best video and audio capabilities, and we warmly welcome CEVA as a strong and trusted partner."

ASUN Technology's second-generation Cupola360 multi-image stitching video and audio SoC AST1230 and its applications are designed for high-performance video conferencing applications. Targeting advanced smart cameras for panoramic video conferencing applications, the AST1230 SoC features a powerful image signal processor (ISP) and intelligent layout processing engine. Combined with ASUN Technology's unique Hyper-stitching® technology, the AST1230 SoC provides superior processing power to enable real-time image stitching on the device. The AST1230 SoC uses the advanced programmable CEVA-BX1 DSP to handle the most challenging audio/voice use cases, providing excellent energy efficiency, making it an ideal choice for a variety of emerging IoT applications. It also supports 8 digital MIC inputs, with powerful audio processing algorithms, including other functions such as far-field voice, beamforming, automatic gain control (AGC), noise suppression, and echo cancellation.

Moshe Sheier, vice president of marketing at CEVA , added: "Video conferencing plays an increasingly important role in most organizations and work-from-home scenarios today, and its related equipment is key to enabling an easy and efficient use experience. We are very pleased to work with leading vendor ASIA to ensure that these conferencing systems provide a seamless and robust voice experience in any user environment. CEVA brings unique value to customers through a hardware-and-software-over-IP approach, and the value-added ClearVox voice software is a great example of this. We look forward to working with ASIA and its customers to bring the next generation of intelligent conferencing systems to the market."

CEVA-BX1 processor Combines efficient DSP computing power with the programming efficiency and code compactness required for embedded applications. CEVA-BX1 uses an 11-stage pipeline and 4-way VLIW micro-architecture, and provides excellent parallel processing capabilities through a single instruction multiple data (SIMD) instruction set, which is widely used in neural network reasoning, noise reduction and echo cancellation, and high-precision sensor fusion algorithms. CEVA-BX1 comes with a complete software development tool chain, including an advanced LLVM compiler, an Eclipse-based debugger, DSP and neural network computing libraries, Tensorflow Lite Micro neural network framework support, and an industry-leading real-time operating system (RTOS) can be selected.

CEVA developed the ClearVox audio front end, leveraging the company's extensive expertise in audio and speech processing, using advanced algorithms to cope with different acoustic scenarios and microphone configurations, including optimized software for speaker arrival direction, multi-microphone beamforming, noise suppression and echo cancellation, as well as related firmware and driver software.