Hewlett-Packard (News - Alert) (HP), the well-known PC manufacturer and provider of hardware, software and services for the IT industry, has tried to make itself a key supplier of cloud computing and big data processing. To better serve its customers and gain a competitive edge in today's market, HP rolled out a number of notebooks and desktops, including a new 27-inch all-in-one workstation and a new Android (News - Alert)-based all-in-one workstation with Touch Capability, at the recent CES 2014.
With this vast offering, HP is attempting to surpass its competition, regain its role of world's largest technology vendor in terms of sales, and move past a period of performance and sales declines. The tech company is, at present, seeking stability with a plan to re-nominate its entire board of directors to lead the company's strategy and corporate development efforts in order to get back on track.
HP has also been making strides recently in its patent applications, and has issued patent requests for other technologies. In addition to its imaging and printing technology, most recently it launched a method for lip activity detection: this technology can really bring speech recognition to the next level. In fact, it allows a device to differentiate between legitimate speech by the user from ambient background noise. “The application describes a system that uses image capture and processing to determine lip activity in response to sounds recorded by the speech recognition system.” The company says the featured patent application with software can basically read lips, as per a recent IPWatchdog post on its website.
HP’s software is able to detect those that speak through a microphone; then, it can convert that response into digital data and into “sounds recorded by the speech recognition system [while able] to identify the background ambient noise and prevent that from affecting a user’s vocal response.”
A patent application has been filed by HP with the United States Patent Office. The software is designed for human speech in computerized recognition systems, ranging from desktop computers to handheld mobile devices.
The patent application almost sounds impossible or futuristic, yet “discloses a novel way of using image processing to aid in speech recognition services,” IPWatchdog explained.
Unlike other USPTO Patents in Speech Recognition and Synthesis, Hewlett-Packard’s intriguing patent application discloses a new way of using image processing – in this case lip reading – to aid in speech analytics. Time will tell what other applications this technology will be able to be applied to.
Edited by Rory J. Thompson