One of the most popular applications in both the consumer and commercial arenas is voice command recognition. Those Alexa And Siri devices that people speak to as well as television remotes that take voice commends have caught on big time. That’s not to mention the automated customer-service representatives that we speak to when we call banks, utility companies, internet/phone/CATV providers, and others that never seem to have an option that fits the purpose of your call.
These applications are constantly being upgraded for better recognition, speed, and, you guessed it, lower power consumption. Voice-activated devices are numbering in the millions, soon to be billions and more. Then figure that nearly all are being accessed countless times 24/7, that’s a lot of power being consumed. Something needs to be done, but what? Easy answer.
At Sensors Expo & Conference 2019 in San Jose, CA, Wednesday, June 26, from 2:30 pm to 3:20 pm PST, one can learn to tackle this issue by attending the session titled, “Machine Learning for Low Power Always-On Voice Control for Battery-Operated Devices” presented by Claus Stetter for acoustic-sensor company Vesper. This session presents a new level of energy efficiency for always-on keyword detect command recognition.
Claus Stetter is the VP of Marketing at Vesper. He joined Vesper in 2017 with 20+ years of experience in the sensor and communication integrated circuit market and 10+ years of building and managing sales teams. Most recently, Claus spent over seven years at Bosch Sensortec, where he managed the entire North American sales and field applications team for MEMS sensors and microphones across key tier-1 OEM accounts.
Prior to Bosch Sensortec, Claus was director APAC Sales at NetLogic Microsystems, a fabless semiconductor company that Broadcom Corporation later acquired. He also held senior sales and marketing positions at Vitesse Semiconductor, Broadcom Corporation, and Allayer. Claus holds a masters in electrical engineering from Hochschule Reutlingen, Germany and has broad expertise in sales and marketing for MEMS microphones and sensors like accelerometers, pressure sensors, eCompasses, IMUs, and gyroscopes, as well as networking and telecom semiconductors. Of note, Vesper is a 2017 Best of Sensors Award winner.
As you can see, Vesper is a regular at Sensors Expo. Mr. Stetter states, “This will be Vesper’s third consecutive attendance at Sensor Expo as an exhibitor. The reason we keep coming back is that it is the only show on the West Coast with a strong focus on hardware and components. The target audience overlaps nearly 100% with Vesper’s customer base, unlike the big shows like Electronica or CES. We have been able to get the word out on our company and make meaningful introductions to new customers and partners.”
Enthusiastic about his presentation, he informs, “Machine learning for low power always-on voice control for battery-operated devices. As the voice interface becomes the next wave in HMI, there is more and more need for mobile, battery-powered devices with such an interface. Battery life is a critical parameter for the adoption of such devices and ETA Compute together with Vesper can offer a dramatic reduction in power which results in better battery life.” The most important concept Claus wants his attendees to leave with is, “A voice interface no longer needs to be relegated to push-to-talk in battery powered applications. Always-listening is now possible without having to take a hit on battery life.”
Additionally, “Vesper will be exhibiting it’s revolutionary piezoelectric MEMS microphone solutions at booth 733.”
Conclusion: if you are involved in machine learning, artificial intelligence, HMI, communications, networking, and more, it will be in your best interest to attend this session. Of course, to do so, you need to do that easy two-step dance. Step one, register for Sensors Expo, and step two, attend the “Machine Learning for Low Power Always-On Voice Control for Battery-Operated Devices” presented by Claus Stetter and Vesper. You’re still here?