Teaching Automobiles to Hear

Makes sense to add another sense to the car.

Teaching Tomorrow's Automobiles to Hear
Fraunhofer-Gesellschaft
Christian Colmer  in ACM

Researchers at Germany’s Fraunhofer Institute for Digital Media Technology (IDMT) have developed a system that allows self-driving cars to recognize external noises such as sirens. The team trained the system to recognize the acoustic signature of specific sound events, using machine learning methods that rely on acoustic libraries. In addition, the researchers created beamforming algorithms that enable the system to dynamically locate moving sound sources. The researchers also developed artificial intelligence-based algorithms to distinguish specific noises the system is designed to identify from other noise. Said IDMT's Danilo Hollosi, “Such systems would be able to immediately recognize the siren of an approaching emergency vehicle, for example, so that the autonomous vehicle would then know to move over to one side of the highway and form an access lane for the rescue services." .... '

Smart Hearing Houses

Virtual Assistants in the smart home are attentive in their hearing.  Why not have the entire house create a hearing and sensing network?  This could be done my connecting existing sensory devices, or built into the house itself, or contained in home appliances.  Here is an interesting play.   Following In Bloomberg: 

Building a House That Can Hear

Audio Analytic, a startup in Cambridge, England, has made it possible for home appliances to hear windows breaking and babies crying.  ... 

As artificial intelligence software continues to advance at an alarming rate, it keeps popping up in new places. A startup based in Cambridge, England, called Audio Analytic, for example, has found a way to inject AI code into smart home devices such as the Amazon Echo and Nest’s smoke alarms. The software can detect the sound of a window breaking or a baby crying and notify the homeowner about the situation.  ... " 

Wireless Earpods

When I think of wearables I do not normally think of earpods, but here is a piece that claims knowledge of IOS and future iPhones possibly including new kinds of earpods that will enhance your soundscape in some new and even profound ways.   Potential wireless 'hearing aid' applications, but also being useful for people with normal hearing.  Unclear how certain any of this information is, but worth following.

More Hearable AI: Mycroft

And how about Mycroft?  There seems to be a market for hearable/vocal channels to knowledge. and cognitive value. In the enterprise, such systems are likely to be attentive and advisory. I recall an example where a group of corporate experts were brought together to advise the CEO in an emergency. Challenge: What would a 'cog' or 'AI' that would also be invited to such a meeting look like? Capabilities?

Ubiquitous Voice Weekly

Just discovered, Ahmed Bouzid, @Didou publishes the Ubiquitous Voice Weekly: News on Voice Interfaces, from Siri to the Amazon Echo  Nicely done about a topic I am passionate about.   Following now as a resource.   Thanks Ahmed, stay on touch.

Google Improves Voice Search

Voice interaction continues to be of interest.  Another channel of interaction that can produce more personal engagement.  Accuracy is important.

In CWorld: Google said it has built a better neural network that is making its voice search work faster and better in noisy environments.

"We are happy to announce that our new acoustic models are now used for voice searches and commands in the Google app (on Android and iOS), and for dictation on Android devices," Google's Speech Team wrote in a blog post on Thursday. "In addition to requiring much lower computational resources, the new models are more accurate, robust to noise, and faster to respond to voice search queries." ... ' 

Talking to the IOT

Intriguing thinking about what a conversation means.  It is not only about context, but building sets of information based on previous questions, answers and reactions.  Expectations about interaction, correctness, trust and risk and even assumed relationships.

In O'Reilly:   " ... When our stuff speaks to us, we exchange more than ideas. We're in the midst of a significant shift in voice and conversation technology, and it's starting to be used in interesting and unexpected ways. ...  "

Forbes on Voice

In Forbes:   On voice interpretation.  Voice to text.  And its use for devices and online.  Not everywhere, since we don't always want to be heard.  They say: " ...  It won’t be long before you wonder how you ever typed much of anything longer than an emoji into your phone–or how, as recently as 2015, your microwave or your clock radio couldn’t understand you. ... ".   In my own case this has been brought up in my experiments with Amazon Echo.  I am now much more willing to speak to a device than I ever was.

Hearables as Advisory Systems

Had not heard the term 'Hearable' used,  but is a mobile device, integrated with an advisory system, linked with sound only to the user.  How much will people normally accept a sound only advisory interface?  Akin to Siri or Amazon Echo, but mobile.   In Retailwire:

" ... Wearable technologies such as Apple Watch, Google Glass and fitness trackers are already collecting information that ultimately may be used to make the wearers' lives easier. But "hearables" — delivering audio cues to users and accepting voice-activated commands — are set to have a huge impact on the Internet of Things (IoT), likely delivering all sorts of new capabilities, even voice-activated shopping, to consumers.

Hearable technologies are projected to rack up $5 billion in sales by 2018, according UK-based Wifore Consulting, or about the same amount as the entire wearables market will take in this year. Why? They tend to be more discreet than other wearable interfaces, and people are already accustomed to in-ear technologies.  ... " 

Amazon Echo as a Task Channel

Have ordered an Amazon Echo.  This is the Amazon always-on microphone/speaker.  Which claims an AI voice recognition and task enablement system.    It is Beta, but a friend who is using it now says it works fairly well.  Similar to an always on and accessible Siri or Cortana system.   It is also being updated as new services are added.

We tested a similar concept in the smart home, specifically a grocery list system that allowed you to readily add list items.   Voice recognition has improved, but is not yet perfect.  Thinking of ways this might be used as a multi tasking channel in the home and the office.  Imagine it being embedded in all rooms, or in an office wall.  Imagine it as a task channel.   Any idea on that for a collaborative investigation?