Emerging technologies


Emerging technologies: voice controlled digital signage


Google Assistant, Amazon Alexa and the like have caught the imagination of home users for devices in the Smart speaker cat- egory, home lighting, heating and so on but pro AV applications are still very much in their infancy. Aware that some early adopters are conducting trials, AV News looks at developments coming around the corner in voice con- trolled digital signage.


Digital signage is developing at breakneck speed, with new display standards, hardware, integration with beacons and other sensors, it really comes as no surprise that the industry isn’t rushing into a new tech- nology before the roll out of ex- isting options is anywhere near complete.


Bur such is the entrepreneur- ial and innovative nature of this industry, there are nearly always new entrants that would like to disrupt the established market by striding ahead with a new technology or price point. Take, as an example, ScreenCloud: this is a 2014 start-up created out of a pro- totype developed at Codegent, the digital product studio:


a host of consumer devices in- cluding Google Chromebit, Am- azon Firestick, Google Chrome- box, Amazon TV, Intel Compute Stick, Chromebase, Chromebox Commercial, Chromebox Mini, Android Box, Intel NUC and others, including some Apple devices."


CE support


ScreenCloud apps run on de- vices as compact as the Intel Compute Stick, the power of an Intel CPU in a small format. These HDMI sticks run the- Chrome App on either Windows or Linux. The Chromebox Com- mercial is a commercial grade device running Chrome OS and is said to be ideal for powering digital displays and vertical ap- plications across a wide range


tell ScreenCloud to show Trel- lo,” perhaps in a corporate ap- plication.


Disadvantages


Naturally, there are numerous privacy


implications, and in- stances of inappropriate us- age to consider - the addition of voice control is tantamount to leaving an unattended key- board and mouse hooked up to the screen. Voice recogni- tion software turns speech into text - as talk into the system and it transcribes what you say - useful for people with visual impairments and for those with physical problems that make typing difficult.


Against that, disadvantages include misunderstanding and misinterpretation. In normal


Web triggers


Despite these drawbacks, many industry pundits are convinced that voice control is imminent, and even established developers are unwilling to leave the field open to innovative start-ups to steel a march.


Marc Benson, CTO at Signage- live said: “We are always looking for cool new ways to control and influence content on screens at Signagelive, in particular us- ing our Web Triggers API so I couldn’t miss this opportunity to experiment with Google Home (others coming soon).” “Using Google Home, the user can start a conversation with our bot and control the content displayed on a Signagelive man- aged smart display or player, in this demo we’ve used a Google Chromebox, however, this would work equally well with any of our supported media players.” “Applications are many and varied from a business assis- tant, through to a retail assistant where products and availability can be shown just by asking for what you are looking for.” In a demonstration, available


MATRIX Voice can be run on a Raspberry Pi or standalone, thanks to an optional module (ESP32) that equips it with a mi- cro-controller as well as Wi-Fi and Bluetooth connectivity.


“We were frustrated by how hard it was to push content to the screens in our office. This was a problem shared by many of our corporate clients as well. What started as a coding hack grew into our biggest product. ScreenCloud was incorpo- rated in April 2015 when we ring-fenced a dedicated team


AdMobilize co-founder and CEO Rodolfo Saccoman: “We believe that voice engagement technologies will make digital signage a more compelling and stickier communications solution for an even broader range of vertical markets.”


of industries. The Chromebase is a powerful, slim, all in one touchscreen device running Chrome OS, and ideal for inter- active displays. With this wide range of for- mats, ScreenCloud has iden- tified applications from edu- cation to hospitality, and the


from Codegent's design and


development resource pool and raised investment. We are based in London, UK and Bangkok, Thailand.”


The ScreenCloud philoso- phy is that playing content on your screen is simple, but users need to make sure they have the right hardware first. “If you have an Android TV then you’re ready to get going. If not, then we can help you choose the perfect device,from


P16 AV News September 2018


close association between the CE host devices supports new voice control technologies de- veloped by the CE manufactur- ers. ScreenCloud has released a video [https://www.youtube. com/watch?v=sW519xtNLTA] showing ScreenCloud digital signage working in conjunction with Google Assistant: “OK Google, show weather in San Francisco”. Google replies ap- propriately and show the weath- er report on screen. Or: “Alexa,


conversation When you talk to people, they decode what you say and give it a contextual meaning. Voice recognition software can do this but may not be capable of choosing the correct meaning. For example, it cannot always differenti- ate between homonyms, such as "their" and "there." It may also have problems with slang, technical words and acronyms. Another advantage often cit- ed for voice control is that it allows the user to multi-task, using their hands for a sepa- rate task. But, you'll have to factor in time to review and edit to correct errors in your spoken communication. Some programs adapt to your voice and speech patterns over time; this may slow down your work- flow until the program is up to speed. If you talk too fast or indistinctly, you'll increase spelling and grammar errors. Getting used to using a sys- tem's commands and speaking punctuation out loud is not al- ways easy. Voice recognition systems fre- quently have problems with ac- cents. Even though some may learn to decode your speech over time, you have to learn to talk consistently and clearly at all times to minimise errors. In digital signage applications, in public spaces, background noise is a big issue. Speech recognition software may not be able to differentiate be- tween speech, other voices and ambient noise. Close, directional microphones and noise-cancelling headsets can be helpful.


on YouTube at [https://signage- live.com/blog/2017/03/23/ voice-activated-digital-sig- nage-from-signagelive-and-goo- gle/] Marc uses voice control to start a Signagelive session, interrogate Google and start a hangout with CEO Jason Cre- mins.


Latest developments


At DSE 2018, AdMobilize in- troduced the MATRIX Voice dev board:


“Put simply, the company that introduced AI-powered audi- ence analytics to the


digital


signage industry is now bringing voice recognition functionality to both manufacturers and sys- tems integrators, alike, through its MATRIX product line,” said AdMobilize co-founder and CEO Rodolfo Saccoman. “We believe that voice engagement technol- ogies will make digital signage a more compelling and stickier communications solution for an even broader range of verti- cal markets. The combination of audience analytics and voice recognition functionality truly represents the next chapter in this constantly evolving indus- try — and AdMobilize is at the forefront of making this chapter a reality.” Available for $55.00, MATRIX


Voice will integrate with any voice recognition service (Am- azon Alexa, Google Assistant or any other third-party service) at any time. “MATRIX Voice puts the power of flexibility directly in the hands of the manufactur- er and systems integrator, free- ing them from being confined to any one of the currently avail- able voice services. This enables plug-and-play custom voice solutions to go mainstream,” Saccoman emphasised.


Equally important, according to Saccoman, is the complete security that MATRIX Voice provides. For retailers who do not want Amazon listening to, recording and storing all store information on its cloud, MA- TRIX Voice is the ideal solu- tion as it provides the capabil- ity to process voice recognition at the edge and not strictly in the cloud. MATRIX Voice can be run on a Raspberry Pi or standalone, thanks to an optional module (ESP32) that equips it with a micro-controller as well as Wi-Fi and Bluetooth connec- tivity. MATRIX Voice has an 8-microphone array, a 3.5 mm audio output jack, 2 speaker outputs, a 3 A audio amplifier, 24 expansion GPIO ports, 64 MB of RAM, 64 MB of Flash, and the second tier Spartan 6 FPGA that allows manufactur- ers and integrators to custom- ize the functions of the GPIO pins as well as implement one’s own audio and voice al- gorithms.


Software


From a software perspective, the company has created 3 library layers to program the MATRIX Voice.


The first lay-


er, HAL, allows integrators to program it in C++, providing the closest access to the hard- ware. The second layer, MA- TRIX Core, contains protocol buffers and ZeroMQ that en- able designers to program the MATRIX Voice in over 40 lan- guages for complete interop- erability with any pre-existing code base. Finally, the third and highest layer, MOS (MA- TRIX Open System), empow- ers users to easily and quickly program the MATRIX Voice in JavaScript with as little as 1 line of code as well as take advantage of the company’s remade infrastructure which includes dashboard, remote deployment through its CLI tool, simple communication between devices (crosstalk), and much more. “Let’s say you have 50 digital screen directories already in- stalled within a mall,” Sacco- man explained. “With MATRIX Voice you can now place it within the top or bottom of the directory (depending on space and optimal audio clarity for the microphones) and then use a 3rd party voice recogni- tion service to create custom ‘wake’ words and responses to enable the shopper to ask, ‘Directory, where is Macy’s?’ and have the directory respond with the route to Macy’s from the shoppers location. All sorts of voice activated cues and information can now be programmed into a digital sig- nage network, opening up the door to a whole new series of applications.”


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20