CEVA audio / voice DSP and software awarded by Novatek for smart TV SoCs

CEVA announces that Novatek Microelectronics, the industry's leading fabless chip design company, has obtained a license to deploy the use of CEVA-X2 audio DSP, ClearVox in its smart TV system-on-chip (SoC) family of products that support multiple microphones ™ voice front-end software and WhisPro ™ voice recognition software, enabling far-field voice wake-up and control functions that are always online Alwaye-On. Novatek is specialized in the design, development and sales of various display driver ICs and SoC solutions.

Moshe Sheier, CEVA's Vice President of Marketing, said: "We are pleased to expand our partnership with Novatek to introduce a powerful voice user interface to its smart TV SoC. Our CEVA-X2 DSP is complementary to ClearVox and WhisPro voice software packages. Let Novatek add new capabilities to create a unique, tailored, multilingual voice experience for its customers. "

CEVA's ClearVox developed in-house leverages its extensive expertise in audio and speech processing, combining advanced algorithms to address different acoustic scenarios and microphone configurations, including for speaker arrival directions, multiple microphone beamforming, and noise suppression Optimized software for acoustic echo cancellation, and related firmware and driver software. CEVA's WhisPro and ClearVox work together to provide customers such as Novatek with a powerful CEVA DSP-based speech recognition solution, enabling them to always listen to smart TVs, smart speakers, smartphones, and Bluetooth earbuds, in order to work with cloud-based voice assistants Services interact. This solution is the overall integration of voice preprocessing and neural network algorithms.It can run locally on edge devices even in noisy environments and far-field use cases, providing a high recognition rate, thereby protecting user privacy and providing minimal delay. CEVA's ClearVox and WhisPro are optimized for the CEVA-X2 high-performance audio / voice DSP designed for intensive audio applications in the smart home, mobile and automotive markets. The CEVA-X2 DSP uses a 5-way VLIW microarchitecture and provides parallel processing through a dual scalar calculation engine, supports single instruction multiple data (SIMD), and optional floating-point units for high-precision algorithms.