Renesas and Syntiant develop a multimodal solution with AI and voice control

Renesas Electronics, renowned for its advanced semiconductor solutions, and Syntiant, a deep learning chip for intelligent voice and sensor processing in edge devices, announced the joint development of a multimodal solution with AI and voice control. It allegedly enables contactless interaction and image processing in artificial intelligence-based IoT systems and edge systems such as self-service kiosks, security cameras and video conferencing systems, as well as smart devices such as cleaning robots.

The solution includes the AI-powered Renesas RZ / V microprocessor and the Syntiant NDP120 Neural Decision Processor for voice and imaging with low power consumption. These components work together to ensure continuous operation, with fast voice activation from standby mode to perform object recognition, face recognition, and other vision-related tasks that are critical functions in security cameras and other systems. For example, while user-defined voice prompts control system activation and operation, AI recognition monitors operator behavior and monitors operation or alerts when suspicious activity is detected.

This solution is called multimodal because it allows you to use two channels – visual and voice. This architecture simplifies the creation of contactless user interfaces for artificial intelligence systems. The use of a dedicated energy efficient chip for voice recognition reduces standby power consumption and speeds up system development as software can be developed independently of the AI-related functionality for machine vision.

A reference design for the new multimodal solution is already available, including schematic diagrams and component nomenclature.