“OK Google” has changed the world as we know it, or at least the Android world as we know it. Either way, true
fanboys believers know that Google’s voice recognition software is the most robust in the industry and the most capable of understanding regular, colloquial speech.
As voice search continues to evolve, its ability to integrate with apps has also improved. Today, Google announced a further step towards that integration. Third-party developers can now employ custom actions that integrate with voice search without the need to open the app first. Google has roped in a specific set of partners to develop this newfound synergy. And it’s all about synergy, folks. Who these partners are hasn’t been disclosed, but the gif below highlights both NPR and TripAdvisor, so let’s not assume that Google woke up today feeling magnanimous and in the mood to offer free ad space to both parties just for funsies.
Has Google Now ever misheard a word you meant to say?
Chances are, it has. Although the service is pretty damn accurate, it’s still a developing technology and has a ways to go before being perfect. Google knows its product isn’t perfect, and in addition to improving voice recognition in Google Search, it has now added a “no I said ___” command for when the app mishears something you said.
If you didn’t know already, the Nexus 5 actually has a low-power audio processor for “always-on” listening (OK, Google) functionality. However, due to the limited functionality of the feature and its closed-source nature, it only works on the home screen and not when the device is asleep.
However, developer Guillaume Lesniak from the OmniROM team has managed to access the dedicated low-power audio core on the Snapdragon 800 chip to make use of this feature. The functionality is only a proof of concept for now, and although we could see it in later ROMs, it is unlikely because of Qualcomm’s re-distribution licensing issues. Hit the break for a video of the feature.
Here’s the drawback to Google recently implemented voice recognition into Chrome; malicious websites can utilize that voice recognition to listen in and possibly record you.
Before you freak out, that sounds significantly worse than it actually is. The “exploit,” according to developer Tal Ater, involves a website asking for your permission to use your microphone for whatever purpose. Afterwards, that site can exploit a bug in Chrome’s voice recognition to listen in on you. A site may launch a pop-up to continue listening in even if you’ve closed the tab for that particular site. Read more
Have a dog and ever uttered the words, “what are you barking at?” HTC may be coming to your rescue. Yesterday HTC posted a photo on their Facebook page showing a “dog translator” app that will let you figure out just what your dog is trying to tell you.
OK, maybe not really. Despite reports floating around the Internet about this development, TalkAndroid did not fall for it. It appears enough people were taken by the idea that HTC took action to officially state that no such dog translator was in the works. The photo that was posted was some creative work from HTC as they mused about the possibility of “voice” recognition software that could be applied to our canine friends.
Perhaps their developers have been working too hard in response to the voice recognition wave sweeping the smartphone market? Or possibly HTC just wanted to have a little lighthearted humor to offset the news about HTC pulling out of Brazil?
Whatever the reason, it’s refreshing to see a company have some fun on a day other than April Fools’ Day.
source: HTC on Facebook
As our smartphones become well, smarter— there’s a need for the smartphone technology to evolve as well. Among the cooler features of most smartphones is voice and speech recognition. Some of the most recognizable speech and recognition companions include Apple’s Siri, Cluzee or Samsung’s upcoming S-Voice. While voice and speech recognition is part of the norm among smartphone users, most people aren’t concerned or interested with how the voice and speech utilities work because hey— people know it’s complicated and really just care about the utilities functioning correctly. That’s why it’s up to companies like Sensory Inc. to show the average consumer how really complicated features are made into simple products like recognizable speech and voice utilities. The underlying goal for Sensory Inc. is to simply make voice recognition smarter. The company specializes in voice recognition, speech recognition and music synthesis— and has had its technology featured in Samsung smartphones, car stereos and even in NASA’s Mars Polar Lander.
Sensory is trying to revolutionize voice and speech recognition by creating TrulyHandsfree, which looks to evolve our interactions with our smart devices. Ponder this: we have our various devices from televisions to smartphones which recognize our voices through different user profiles, but switching between profiles can be a bit of a hassle— so couldn’t changing between profiles be simpler? Sensory promises every aspect about speech and voice recognition will be simpler with TrulyHandsfree. Essentially devices that use Sensory’s TrulyHandsfree technology will be able to listen to and identify who is speaking to it– similar to how you recognize the voice of someone who is calling you. Moreover, the technology may someday make things like passwords, pattern unlock and fingerprint sensors obsolete. An example of this is a smartphone user saying “Hi Galaxy” to wake or unlock up his or her Galaxy S III smartphone as opposed to having to enter in a passcode lock key or swipe your finger. Neat isn’t it?
TrulyHandsfree is a technology that’s here already and ever-evolving. Sensory claims its technology has already set a high standard— a 95% accuracy rating which is solid, but there’s some room for improvement as you already know. It’s currently featured in Sammy’s S-Voice technology for now— but here’s hoping it will be featured in more and more smartphone in the
years months to come.
source: The Next Web
via: Android Authority
Google’s response to Apple’s voice recognition/virtual personal assistant, previously dubbed “Project Majel” has reportedly gone through a name change, according to TechCrunch. Although Google has had in-house voice recognition for a while, Assistant plans to go above and beyond the current capabilities of Siri. The assistant team of developers is headed by Amit Singhal, the Google search engineer responsible for the search algorithms we use on a daily basis, including the ranking system that Google uses to rank pages as results. Singhal has a fairly impressive resume and is the recipient of plenty of prestigious awards for his service and innovations in the information retrieval sector. With him at the helm, it’s no doubt that Assistant will turn out exactly the way we hope.
According to reports, the project has three parts:
- Get the world’s knowledge into a format a computer can understand
- Create a personalization layer — Experiments like Google +1 and Google+ are Google’s way of gathering data on precisely how people interact with content.
- Build a mobile, voice-centered “Do engine” (‘Assistant’) that’s less about returning search results and more about accomplishing real-life goals.
Unlike Apple and their closed loop of hardware-software interaction, Google plans on in the involvement of the Assistant project and third-party developers, opening up the prospect of tying Google Assistant into the depths of your device. Cool, huh? The project has reportedly been given a Q4 release period, which is a pretty long way down the road, but better late than never, right? Hit the break for an interesting video regarding the evolution of search and its future capabilities.
Ever since Apple released Siri for iOS, Android developers have been coming out with their “Siri-killers”. We’ve seen Speaktoit Assistant, Cluzee, Iris, and even Google’s upcoming Project Majel, which promises to improve the current Voice Actions for Android. The latest is called utter! by XDA member brandall, and promises something a little different.
Through a video the developer posted on YouTube, we find out that utter! knows about and can use the apps already installed on your device. For example, when asked “search eBay Galaxy Nexus,” the app responds by displaying the results through the installed eBay app rather than opening a browser. Of course, if the eBay app was not installed, utter! would indeed use eBay’s web site to return the results.
HTC EVO 4G owners are seeing an update today to correct some minor bugs, one of which includes voice recognition accuracy. The update will take 1 to 2 weeks to complete, and you will be notified when it is pushed to your phone. Alternatively, you can hit Menu/Settings/About Phone/System Updates to see if you can manually pull it. This update will bring your software to version 4.53.651.1
- Integrated voice recognition accuracy with voice search and voice-to-text
- Image size issue with contact transfer over bluetooth
- Images cut off in Picasa
- Group contacts not in alphabetical order
- Out of memory, Unable to send SMS
It’s proven dangerous, but many drivers are guilty of it from one time or another. I’m talking about texting while driving, where a few quick strokes of your touch or keypad can turn deadly.
StartTalking, a new texting application from AdelaVoice, looks to give drivers no excuse for texting while driving, with this hands free, voice controlled solution. The application is completely hands and eyes free, allowing the user to initiate, compose, review, edit, and send text messages entirely by voice command.
The app is currently in open beta for Android 2.0 and above devices, and can be found in the Android Market, or by visiting http://m.starttalking.com from your phone.
Continue reading after the break for the full press release.