NaviScribe covers a set of navigation technologies that combine robotic vision algorithms with appropriate hardware to achieve efficient pose recovery and 3D motion capture. NaviScribe navigation engines and software can be integrated in low-power and small form-factor devices. Custom or consumer devices that already have appropriate on-board sensors (e.g., a CMOS camera and inertials) can take advantage of NaviScribe to recover their true 3D position, orientation and motion parameters. NaviScirbe was developed by Electronic Scripting Products, inc. in Palo Alto, California.
Put a pen or a stick in a little child's hand and you'll never doubt the essential naturalness of handwriting, drawing and gesturing. Nevertheless, the dominance of personal computers has made keyboard users out of all of us. Meanwhile, pure touch interfaces in most smart phones and tablets (e.g., Apple's iPhone and iPad) have pushed us to poking and drawing with fingers. We have even resorted to collecting digital signatures from scribbles made on small touch-screens with one's pinky. Such "marks" rarely resemble true ink signatures. What's more, their biometric value is clearly suspect in verifying financial transactions and thus the process of authorizing purchases by signature has been relegated to a mere formality that nobody takes seriously. That is not only a pity, but may also be an indirect cause of huge losses to financial institutions and insurers. As law enforcement investigates tens of millions of stolen cards and compromised identities it could greatly benefit from accurate digital signature capture that only the owner can reproduce in person and on the spot. This old-fashioned way of identity verification could undermine the basic premises of criminal operations such as the latest slate of Russian identity hacking actions in which over 1.2 billion passwords were stolen (O'Toole, J. & Pagliery, J. "You've been hacked! Here's what to do", CNNTech, Aug. 6, 2014). After all, what are all these funny-sounding passwords worth if at the crucial moment you are being asked to produce a high-quality digital signature in real time and/or a special mark known only to yourself? Meanwhile, the proliferation of pass-phrases including strange symbols that make them hard to remember, as well as the tedious "secret questions" about the name of your high school, color of your hamster or the brand of your first car are proving to be a poor tool in the process of identity verification. Memories fail, disks get erased and papers full of sensitive information are either misplaced, lost, or even worse, stolen along with the cards. Most of us, when confronted with a typical password entry situation simply click the button that says "forgot my password". We then reset it to something just as difficult to remember as the previous one and repeat this operation on the next account login. The fact that newest operating systems promise to keep all of these hard-to-remember character combinations easy to access under a simple password for the user's convenience and store copies "in the cloud" is hardly a reassuring move to be sure.
Projects that attempt to obtain a "verified" signature by having a person type their name on a computer to print it in a signature-like font are likely not the correct answer either. After all, someone who simply takes over the computer terminal of the intended signatory at the right moment can input names and initials just as easily as the intended recipient. Equally troubling is the prevalent practice of accepting as full and legally binding signature the person's name entered by keyboard and enclosed between two forward slashes, as in: /My Name/. Would forgetting the slash while typing in your purported signature characters render your John Hancock invalid? Apparently the U.S. Patent and Trademark Office as well as various other institutions and businesses believe so. These beliefs have created many unexpected and costly consequences to those who imporperly applied this new age "signature". With no on-the-spot biometric verification formerly provided by a user's signature being generated right there and then (verifiable against a signature card kept on file -- as formerly done by all banks) many powerful tools in our counter-fraudulence toolbox are being sacrificed in the name of questionable user convenience.
Now this is changing. Tablets and tablet computers designed to capture writing and sketching on their displays with styluses are increasingly prevalent thanks to Samsung and others. Digital pens that write on paper and simultaneously capture digital ink are being pushed by technology innovators such as Wacom and LiveScribe. These devices are doing well on international markets.
User Interfaces intended for data input and gaming are moving in the direction of capturing three-dimensional (3D) gestures. Starting in 2006, the Nintendo Wii has convinced millions that performing such gestures is more intuitive than pushing buttons and twisting knobs. The sales figures for Microsoft’s Kinect and Sony’s Move are confirming this realization.
Since 2002 the researchers at ESPi have been preparing for a transition of User Interfaces to true 3D. Our 3D input solutions, called NaviScribe, are based on accurate full motion capture. NaviScribe recovers movements of user-manipulated devices in up to six degrees of freedom (x, y, z, pitch, yaw, roll). These six degrees of freedom, collectively referred to as pose or as the 6 DOFs, are expressed in stable world coordinates. Conveniently, the same coordinates also describe the user’s familiar environment.
ESPi’s approaches build on optical pose estimation originally developed in computer vision. Advances in electro-optics and a wise choice of features for optical tracking support efficient, rapid and low-cost pose recovery in stable coordinates (absolute pose). For example, IR LEDs enable absolute pose recovery at frame rates above 100 Hz. Edges of display screens also present high-contrast optical features that support robust absolute pose recovery. In addition, any screen cooperating with NaviScribe can present a suitable pattern to further aid in optical pose recovery.
Some NaviScribe units combine optical pose recovery and relative motion measurements. The latter are obtained from inertial devices, such as gyros and/or accelerometers. Fusion of data supplied by optical and inertial sensors is advantageous in low-cost and high-speed systems. Such systems can support high precision 3D wands and digitizers, SmartTV remotes with direct pointing (remote touch as defined by Philips' uWand technology), true 3D game controllers (think Wii wand that really knows its 3D pose), as well as styluses sufficiently accurate to record handwriting.
ESPi, Inc. is a development-stage R&D company officially founded in 2003 and located in Palo Alto, CA.Since its inception in 2002 and prior to raising the first round of funding from angel investors in 2003, ESPi commenced a concerted drive to secure patents for all innovative aspects of its technical solutions. In continuing this effort without interruptions, ESPi has been able to amass a substantial domestic and international IP portfolio. Our patented technologies derive from novel applications and integration of mechanics, electronics and optics, as well as pose estimation algorithms. At the present time, ESPi is continuing its search for qualified third parties to explore the development of beta units and to support any productization studies by OEMs.
Marek Alboszta, Inventor, CEO
Dale H. Buermann, Inventor
Stewart R. Carl, Inventor
Hector H. Gonzalez-Banos, Ph.D., Inventor, Vision Algorithms Specialist
Timothy J. Hawks, Inventor
Michael J. Mandella, Ph.D., Inventor, CTO
Guanghua G. Zhang, Inventor