Synaptics shows off edge AI kit for IoT and monitoring applications

Synaptics has launched its edge AI evaluation kit (EVK) that employs human recognition, sound, and motion detection capabilities on its Katana system-on-a-chip (SoC) to boost biometrics and other AI applications at the network edge.

The Katana EVK integrates detection of human bodies, sounds, voice, and movement in a low-power edge AI package for IoT applications in smart homes, smart cities, security, buildings, industry, and monitoring needs.

Synaptics quotes a report from Valuates Reports that says the rising demand for real-time, low-latency response edge AI in automobiles and consumer electronics faces the challenge of power consumption and size constraints in edge semiconductors. The company seeks to resolve this problem with its SoC that is low-power and compact while offering sufficient computing power for effective AI and biometrics.

“The market for low-power edge AI solutions is characterized by a broad spectrum of applications that have specific use case requirements,” says Ananda Roy, senior product manager of low-power AI at Synaptics. “What they have in common is that they all require intelligent and efficient detection and analysis of image, audio, voice, and motion data.”

In a video demonstration, Roy shows how the people detection system would function in a low-power situation. A provisional security gate with a camera-equipped Katana EVK dangling atop the frame connects to a Raspberry Pi Pico laptop that projects a video feed. The AI algorithm detects the subject walking into its vision and forms a yellow border around him. Roy joins the camera feed, and the Katana EVK counts two individuals.

Roy mentions specific uses for the Katana EVK such as occupancy and space management in smart buildings, occupancy limit compliance in crowded venues like theaters and stadiums, and eliminating false detections. As a result, it would enhance the user experience and extend the life of batteries and battery part cameras used in home security.

“Edge AI and sensor fusion are complex, multifaceted fields of application development and we know from experience that our customers benefit most by having a more complete platform of hardware, software, wireless connectivity, and technical support to get their IoT concepts to market,” Roy concludes.

Industry-watchers see major potential impact on the biometrics industry from edge computing approaches.

Article Topics

AI  |  AI chips  |  biometrics  |  biometrics at the edge  |  IoT  |  monitoring  |  person detection  |  research and development  |  Synaptics