Related
Is there any app out there that will allow speech to text from within any application, or whenever you pull up the keybord or something? Preferably multi-lingual.
HTC_IME
You can install the modified HTC_IME keyboard that comes with the speech-to text-button on it, you can make a modified version of it at gimpsta. com/themer/index.php. It uses the google speech thingy, so you should be able to speak in more than English I think.
Then to use it, if you go settings > Language & Keyboard, yuo should be able to eck HTC_IME mod. Then long click any text area > Input method and chose HTC_IME mod.
What is REALLY stupid thing by google/android is that they dont offer a text-to-speech / speech-to-text/ voice search to languages that literally pronounce every single letter the way theyre written. It would be the easiest thing in the world to do. This would offer a solution to many languages (such as Finnish thats heavily ignored by google/android in every other aspect too).
Just my 2cents. Ppl at google plz read this.
wardy666 said:
You can install the modified HTC_IME keyboard that comes with the speech-to text-button on it, you can make a modified version of it at gimpsta. com/themer/index.php. It uses the google speech thingy, so you should be able to speak in more than English I think.
Then to use it, if you go settings > Language & Keyboard, yuo should be able to eck HTC_IME mod. Then long click any text area > Input method and chose HTC_IME mod.
Click to expand...
Click to collapse
JuniperFIN said:
What is REALLY stupid thing by google/android is that they dont offer a text-to-speech / speech-to-text/ voice search to languages that literally pronounce every single letter the way theyre written. It would be the easiest thing in the world to do. This would offer a solution to many languages (such as Finnish thats heavily ignored by google/android in every other aspect too).
Just my 2cents. Ppl at google plz read this.
Click to expand...
Click to collapse
Thanks for the replies. First, HTC IME does have speech to text, but only in english. Choose another language and the mic button automatically becomes just a simple settings button. SO much for that...
As for Juniper's rant, I agree. I love Google. I mean, I'm a God damn fanboy! I live and breathe Google and have tried to convert all my friends and family and misc forum users to Google services whenever possible. But the fact that they mainly play for the big boys (English speaking countries) and always try their services in a fixed set of countries first, kinda annoys the hell out of me and makes me feel left out. ****, I even went as far as to buy a Nexus One online for almost double their worth because they didn't go on sale in Portugal for a very long time. I just consider myself lucky I was able to sell it on for a small profit despite the already ridiculously high price I had paid for it to begin with. But even after that I opted for a HTC Desire just because I wanted the best Android experience possible (at the time). I am a Google advocate. ****, Google has probably made (what to them must be a small amount but to normal folks would be) a metric ****load of money just on services bought by people I have recommended use Google over the past decade or so. Me, myself, I use Google for just about all my online needs from shopping to social networking. So no, this issue won't make me want to stop using Google, for sure, but it does make me feel my "relation" with Google is just that little bit less "special". They do tend to be biased and not support the little guys.
That said, I still have no solution to my question. Ideas?
In this Noob series we will look at How to: Install Google Voice Search with Speech Recognition Software Noob (Simple) Edition.
If you have already followed our [HOW TO] Install ICS Google Applications and Google Voice Noob (Simple) Version
then you already have Google Voice Search with Speech Recognition Software. It was included with the google_apps package. You may ignore this tutorial.
Prerequisites:
1. How To: 7" Kindle Fire HD Root - Noob (Simple) Version http://forum.xda-developers.com/showthread.php?t=2069117
Speech Recognition Step 1: Downloaded the correct package here: VoiceSearch
Speech Recognition Step 2: Extract VoiceSearch.zip to VoiceSearch on your Desktop
Speech Recognition Step 3: Copy libvoicesearch.so, UniversalSearch.apk and VoiceSearch.apk from the folder VoiceSearch to the Download folder on your Kindle Internal Drive
Speech Recognition Step 4: Using Root Explorer or ES File Explorer, copy /sdcard/download/libvoicesearch.so to /system/lib/libvoicesearch.so change the libvoicesearch.so permissions to Owner: Read Write Group: Read Others: Read
Speech Recognition Step 5: Using Root Explorer or ES File Explorer, copy /sdcard/UniversalSearch.apk to /system/app change the UniversalSearch.apk permissions to Owner: Read Write Group: Read Others: Read
Speech Recognition Step 6: Using Root Explorer, copy /sdcard/VoiceSearch.apk to /system/app change the VoiceSearch.apk permissions to Owner: Read Write Group: Read Others: Read
Speech Recognition Step 7: Reboot the Kindle
Speech Recognition Step 8: Now go into Apps, you will see a new Icon named Search. Launch Search, then click on the little Microphone, speak clearly (I used the word Test).
Speech Recognition Step 9: The application will next ask you which browser do you want to use to display the results (I use Opera).
Speech Recognition Step 10: Next Open up Google Play Store, Search for "AVC (Alice)", here is the link: https://play.google.com/store/apps/details?id=yourapp24.android.tools.alice_lite&hl=en
Speech Recognition Step 11: Choose to Install AVC (Alice), Google Play Store will then Download and Install AVC (Alice). This is a full blown Speech Recognition Software, it will even have a conversation with you!
Speech Recognition Step 12: After you test the features we highly recommend that you upgrade to AVC (Alice) Pro, here is a link: https://play.google.com/store/apps/details?id=yourapp24.android.tools.aivc
Speech Recognition Step 13: Finally, using Google Play Search for Alarm Clock Plus, here is the link: https://play.google.com/store/apps/details?id=com.vp.alarmClockPlusDock
Speech Recognition Step 14: Choose to install Alarm Clock Plus, Google Play Store will then Download and Install Alarm Clock Plus
Then just like that, your new journey begins!
AVC (Alice) Pro Notes:
Alice is your personal assistant. Alice can help you do things quickly and easily.
With Alice, Android users get a helpful assistant at hand, which understands more than just commands.
You can have a conversation with Alice, give her instructions or ask for general terms of information.
Currently Alice can assist you with following functions:
Call, SMS, Email, Translation, Navigation, Alarm, Timer, App Launcher, Calculator, Facebook, Websearch, Weather, Search images.. and many many more.
You can also define your own commands with Alice and you have the opportunity to control devices which are accessible via a web interface.
Additional to the lite version: Calendar, Play Music / Videos and Receiver Control for Enigma2, Dialog-Mode, Listen in Background.
Alice knows a lot and is constantly learning new things, some examples:
- Do I need Sunglasses?
- How far is it from New York to San Fransisco?
- Director of Inception
- Zap to CNN
- Set up a meeting on monday at 12 o'clock coffee with Megan
- How does my day look on friday?
- Play Whitney Houston - Step by step
- Show me my pictures
- How's the weather?
- Call John Smith
- Send a message to John i'm coming later
- Wake me up in half an hour
- What is 5 + 7 * 3
- How many contacts do I have?
- How many songs do I have?
- Set Timer for 20 seconds
- What's the definition of aspirin
- Translate into german hello how are you
- Start WhatsApp
- What is 5 meters in miles
- Do you know Barack Obama
- Navigate me to the next McDonald's
- Make the light on
- Show me New York on the map
- How many contacts do i have?
- Post on Facebook what's up?
- Search for pizzeria in New York
- Who is Angela Merkel?
- Mute / Silent
- Navigate to New York
- What is the capital of Germany
- How many people live in Berlin?
- How old is Rihanna?
- When is her birthday?
- Show me a picture of Michael Jackson
- etc...
Many options! Design, voice, App2SD and more. Button Animation with Alpha!
Like us on facebook: http://www.facebook.com/YourApp24
Requirements: Google voice search (available in android market for free) or Vlingo recognition
Currently supported languages are German and English.
If you have problems with setting the alarm, you can download Alarm Clock Plus in the Market for free.
The Online-TTS is powered by iSpeech®
How do I use text-to-speech (“read to me”) feature on Kindle?
Kindle can read text to you. In books and periodicals where text-to-speech is enabled (publishers may opt-out of the feature) you can start, stop and pause either by going to the text settings screen by pressing “Aa” button or by using the hotkeys: Shift-Sym to start/stop and once it’s running – spacebar to pause/resume. In the text settings screen you can also configure speech speed and switch between male and female voice. To start reading from specific location, select the location with 5-way controller and start text-to-speech.
Awesome man, can text and call as well with GV?
Sent from my ADR6400L using xda premium
Google Voice Search
DNak206 said:
Awesome man, can text and call as well with GV?
Sent from my ADR6400L using xda premium
Click to expand...
Click to collapse
That depends if the application calls to the VoiceSearch.apk module. Yes, it should work if the application was designed around the GV Module. Since we left everything intact, there is nothing to stop another application from calling and using the VoiceSearch.apk module. This is where people like you come in: Install it, then test applications that were design to work with Google Voice and reply back with the results.
If the application was not designed around the VoiceSearch.apk, there is little chance it will work. So, if you have a friend with an Android, ask them what applications use Google Voice. For example, I know Google Voice works with the VoiceSearch.apk module when you search in Google Voice.
Keep me posted, I will be interested in your feedback.
DNak206 said:
Awesome man, can text and call as well with GV?
Sent from my ADR6400L using xda premium
Click to expand...
Click to collapse
I believe there's confusion around what is being talked about here... This thread will help install voice enabled google search. Google Voice is a completely different product. See: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&sqi=2&ved=0CDQQFjAA&url=https%3A%2F%2Fplay.google.com%2Fstore%2Fapps%2Fdetails%3Fid%3Dcom.google.android.apps.googlevoice%26hl%3Den&ei=s7zeUIyOHuSviAKQj4DgBg&usg=AFQjCNFBqz0CSZ3-2SmgeVYgLuK6o5DyBA&sig2=AVh3iLZEvXgi9ik3LwPd2g
Google Voice Search
helix4u said:
I believe there's confusion around what is being talked about here... This thread will help install voice enabled google search. Google Voice is a completely different product. See: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&sqi=2&ved=0CDQQFjAA&url=https%3A%2F%2Fplay.google.com%2Fstore%2Fapps%2Fdetails%3Fid%3Dcom.google.android.apps.googlevoice%26hl%3Den&ei=s7zeUIyOHuSviAKQj4DgBg&usg=AFQjCNFBqz0CSZ3-2SmgeVYgLuK6o5DyBA&sig2=AVh3iLZEvXgi9ik3LwPd2g
Click to expand...
Click to collapse
You are correct, there can be a bit of confusion. In our How to: Install
ICS Google Applications with Google Voice Search Noob (Easy) Version
we actually use the Talk.apk which Google Voice. The problem is that we
have not found a way (yet), to pip this over to applications like Voice Chatting. Although we are
working on it and should have this functional by February 2013. Either way, this step will be
a Prerequisite in getting the entire package to work properly. Start here, give us feedback.
Speech Recognition Software
To All,
We just ported over, AVC Alice, which is used like an assistant.
We will be adding the install notes in the next 30 minutes.
great!
Great Tutorial. I've got it installed and it works great. It crashes on my KFTT when I ask it to make a note for me. It tries to attach a file and fails to find an app I think.
AVIC (Artificial Intelligent Voice Control) (Alice) - Missing App
bbanghyung said:
Great Tutorial. I've got it installed and it works great. It crashes on my KFTT when I ask it to make a note for me. It tries to attach a file and fails to find an app I think.
Click to expand...
Click to collapse
bbanghyung,
Wow, I am sorry. I know I responded to this Thread 8 plus hours ago but my response never made it
into the XDA Thread. Hmmmm. Anyway, you nailed it on the head! The problem when you tell AVC Alice
to "Open Note", she should respond with her intended action. Now what she wants depends on the Launcher used.
In my case, I use Go Launcher EX and HD, she detects this and knows to call up "Go Note Widget", so this
needs to be installed and setup. Just take your time, go into Settings, you will see there is a lot to AVC.
So far how do you like AVC? Fun stuff and really helpful once it has been setup.
Anyone find a way to access the voice in/out settings? I have a couple of Ivona TTS voices I'd like to try on Alice.
TTS EXTENDED looked promising, but it either forced closed or froze. There might be a trick to it. Anyone?
Alice is great. I still can't get notes to work, it just says an error has occurred. I'll take a look at it again later though. Also alarm clock plus crashes every time I attempt to access it's settings. I have moved it to my system folder and restarted and still same problem occurs.
Sent from my KFTT using xda app-developers app
All Alice is giving me today is "an error has occurred." And just like a woman, she won't tell me what's wrong. And I apologised and everything...
Edit- It may have been the FB reader tts (deleted) or just low memory (I deleted some deadwood), but Alice & I seem to be on speaking terms again. Still hoping someone knows a way to access the tts-picking menus. I'd love to experiment w/voices. I tried several things last night, but no joy.
AVIC (Artificial Intelligent Voice Control) (Alice) - Settings
BELASCO said:
All Alice is giving me today is "an error has occurred." And just like a woman, she won't tell me what's wrong. And I apologised and everything...
Edit- It may have been the FB reader tts (deleted) or just low memory (I deleted some deadwood), but Alice & I seem to be on speaking terms again. Still hoping someone knows a way to access the tts-picking menus. I'd love to experiment w/voices. I tried several things last night, but no joy.
Click to expand...
Click to collapse
BELASCO,
Yes AVIC Alice requires her space (memory) and if you do not give her space, she can become
snappy, kind of like PMS I guess. The message "an error has occurred" can be memory issues,
missing application (like if you say "Open Note", she requires Go Note), application conflict, etc.
We included this package since it is by far the best in the industry, like Dragon Naturally Speaking
is the best for a PC. I an excited to hear you are interested in experimenting with the voices. This
is exactly my goal, to improve the Kindle and spark enough interest to have others involved with
the development. Please send us responses as you work with AVC, we will be really interested
in your work. We have a few developers on site, so send me a reply with the details of your
interests with AVC and I will forward the request to my Senior Developers.
I was looking to change the KFHD's voice (and Alice's) to Ivona's Gwynneth or Amy, but found the usual Android menu lacking on the Kindle. Know of any way to get at it?
Menu Missing
BELASCO said:
I was looking to change the KFHD's voice (and Alice's) to Ivona's Gwynneth or Amy, but found the usual Android menu lacking on the Kindle. Know of any way to get at it?
Click to expand...
Click to collapse
BELASCO,
When you have AVIC Alice opened, look down at the home task bar, between the < arrow and the Magnifying Glass (Search) is a
upside down U with 3 lines through it. When AVC is open, press on this upside down U with 3 lines through it and you should
see a menu pop up. Do you see the menu options - Help - Text Input - Manager - Settings?
The choices see on my KFHD are "male, female & system" in settings.
AVIC (Artificial Intelligent Voice Control) (Alice) - Settings
BELASCO said:
The choices see on my KFHD are "male, female & system" in settings.
Click to expand...
Click to collapse
BELASCO,
Yes, of course, you have the free version which is limited in the options. It is worth it, upgrade to Pro Version
https://play.google.com/store/apps/details?id=yourapp24.android.tools.aivc
it costs a whopping $2.36 and adds an enormous number of features and options, including everything you
want to do. Play with voices, custom responses, etc. Over 100 options added to Pro. The Basic Free
Version is meant for testing and simple development to integrate AVIC Alice into other application, then
it is meant to be upgraded. Do the Upgrade, you will love it.
I HAVE the pro version. What menu items am I missing?
AVIC (Artificial Intelligent Voice Control) (Alice) - Settings
BELASCO said:
I HAVE the pro version. What menu items am I missing?
Click to expand...
Click to collapse
BELASCO,
Go back to Google Play Store, search for AVC (Alice) Pro. Then click on the AVC (Alice) Pro Edition.
Does it show Installed? If yes, uninstall, reboot and reinstall. You are missing a lot:
Help > Help System
Text Input > Place to make Custom Voice Commands
Manager > Place to Manage Commands and Alias's to Contacts
Settings > Here you have about 80 Options inside of Settings
If you click on Settings > About us it will advise you of the Version.
Version 2.8. I did as you suggested reboot it all but it's made no difference in my menu. And it is the pro version. There are only 3 menu choices under voices.
AVIC (Artificial Intelligent Voice Control) (Alice) - Voice Settings
BELASCO said:
Version 2.8. I did as you suggested reboot it all but it's made no difference in my menu. And it is the pro version. There are only 3 menu choices under voices.
Click to expand...
Click to collapse
BELASCO,
I am sorry, I misunderstood you, I thought you were saying the only options
you had were 3. My mistake. You are fine, I am wrong. Maybe if you go
to the developer site http://www.yourapp24.com/ they
may offer other voices. You may also email them at [email protected]
I have avoided Google now for the previous two iterations of phone... but I'm giving it a shot on the S4!
There are two outstanding Google Now questions that I can't seem to find a clear answer to by Googling (there's a joke in there somewhere).
1) How do I actually trigger Google now? I've tried saying Google, that doesn't work. The best I've been able to do is map the double home button to launch voice input, but that's s poor solution. If I have to use button presses to trigger it, then I may as well just do whatever I was trying to do in the first place with my fingers. IE, it's no time time savings to double press the home button and then say a command, vs just clicking on the app icon I'm trying to launch (probably a time penalty in fact).
What am I missing? Why isn't my Google Now listening to me in the background? Am I just using the wrong trigger word?
2) The beep that follows voice recognition is incredibly loud, substantially louder than the speech response that Google feeds back. How do I disable the beep entirely? It's redundant and disruptive. The best solution online seems to be to mute notifications in general. That's a non starter.
You trigger it by starting the Google app and following the instructions. It's the google app, not the google+ app or the google settings app.
After that it just kind of runs. starts out slow at first but more cards appear over time, but you can pull it up anytime by running the google app.
acruxksa said:
You trigger it by starting the Google app and following the instructions. It's the google app, not the google+ app or the google settings app.
After that it just kind of runs. starts out slow at first but more cards appear over time, but you can pull it up anytime by running the google app.
Click to expand...
Click to collapse
Yes.... I realize how to run the app.... and how to follow its instructions
How do I trigger it, that's my question/query/frustration. As mentioned, it works fine if I double-tap the home button to launch the Google Now app manually. It also works fine if I go into my apps and find the Google Now icon, and then launch it that way, or if I use a shortcut, widget, etc to open up the app.
What I can't figure out, is how to actually trigger it during regular use. Short of going in and manually launching Google Now, I can't get it to trigger - I can't get it to listen to me. I've tried saying "Google", "Android, "Ok Android", "Ok Google", etc, and almost every other ridiculous phrase I can think of. Nothing triggers it short of manually launching the app by hand.
Once I set it up, it comes up every time I tap on the google search bar. However, normally I don't mess with it, it just automatically sends me notification cards when it thinks I need them.
For instance right now it's showing me a notification card for the last Sharks Kings game, my upcoming flight and several cards for things I've researched a lot lately like the S4, Voodoo Sound, Vsonic GR02 headphones, Fiio E07K etc. Also has my stocks and the local weather. I didn't "do" anything to get the cards, they just appear in my notification bar or whenever I tap the search bar.
acruxksa said:
Once I set it up, it comes up every time I tap on the google search bar. However, normally I don't mess with it, it just automatically sends me notification cards when it thinks I need them.
Click to expand...
Click to collapse
That's fine - once setup, it comes up every time I tap the Google search bar too. That's not my issue.
My problem is that I cannot get it to trigger by my voice. Unless I manually launch the app first each and every time I want it to do something, it does not respond to my voice. It responds to my voice just fine if I manually I launch the app first, but that's the only time I can get it to work. At all other times, I may as well be speaking to my shoe
Hold menu button for couple seconds.
Easy answer. You can't you have to launch it first.
Sent from my SAMSUNG-SGH-I337 using Tapatalk 2
ripper4209 said:
Easy answer. You can't you have to launch it first.
Sent from my SAMSUNG-SGH-I337 using Tapatalk 2
Click to expand...
Click to collapse
That can't be...? What use would a hands-free system be if you have to use your hands and navigate into an app before it will take hands-free voice commands?
Isn't the whole deal with this voice stuff that you can just talk to it?
In my experience, the hotword detection only works after you've manually triggered the app. It only relieves you from physically tapping the microphone icon.
Sent from my SAMSUNG-SGH-I337 using xda premium
get a bt headset and activate it that way..It can't be 100% hands free; it would have to be on all the time, and then you would complain about how much battery it was eating...
On my Google Now, saying "Google" does work. As long as the actual App has already been activated and is standing by that is.
On my "S Voice", saying, "Hi Galaxy" works, but it is customizable for other triggering phrases.
You are correct about Google Now only responding the first time though. If you want it to voice activate again you have to use the back key to get back to the original screen or completely re-launch it. It just seems that is the way it is.
Wow - this is significantly less impressive than I had envisioned. I was under the mistaken impression that Google Now (and Siri on the iPhone) just responded to their trigger words whenever uttered.
This strikes me as somewhat limited in usefulness until they get these apps to listen constantly. If I need to know the weather, why use my hands to launch Google Now, then ask for the weather, and then wait for it, if I could instead just use my hands to launch my weather app directly?
I must be missing all the excitement - or maybe Siri / Google Now are meant for people who can't figure out their phones?
On the issue of that annoying confirmation sound, I'm halfway there. The APK is in data/apps, and it's called "com.google.android.googlequicksearchbox-1.apk". Inside that apk there is a folder called "res", and inside that there's another folder called "raw". In that folder you'll find the WAV files for each sound.
Unfortunately, that's where my efforts die. I can't seem to open the wav files on a PC (even in audio editing software). I can't just delete them either, or Google Now crashes. I similarly can't just replace them with less annoying wav files from my PC - that crashes Google Now also. I'll see what I can do about finding a way to just upon them for the purpose of lowering their volume (or blanking them) and report back with my results (if any).
rhd-android said:
Wow - this is significantly less impressive than I had envisioned. I was under the mistaken impression that Google Now (and Siri on the iPhone) just responded to their trigger words whenever uttered.
This strikes me as somewhat limited in usefulness until they get these apps to listen constantly. If I need to know the weather, why use my hands to launch Google Now, then ask for the weather, and then wait for it, if I could instead just use my hands to launch my weather app directly?
I must be missing all the excitement - or maybe Siri / Google Now are meant for people who can't figure out their phones?
On the issue of that annoying confirmation sound, I'm halfway there. The APK is in data/apps, and it's called "com.google.android.googlequicksearchbox-1.apk". Inside that apk there is a folder called "res", and inside that there's another folder called "raw". In that folder you'll find the WAV files for each sound.
Unfortunately, that's where my efforts die. I can't seem to open the wav files on a PC (even in audio editing software). I can't just delete them either, or Google Now crashes. I similarly can't just replace them with less annoying wav files from my PC - that crashes Google Now also. I'll see what I can do about finding a way to just upon them for the purpose of lowering their volume (or blanking them) and report back with my results (if any).
Click to expand...
Click to collapse
Don't you think some people DON'T want it on all the time? It could mistakenly trigger it and possibly do an unwanted action. It is fine in it's current implementation. I would think in the future as voice recognition and battery technology improves we might see an always on Google now or siri but right now it would/could cause more problems than create convenience.
Sent from my GT-N7100
A new version of Chrome (for desktop) is coming out that supposedly will do this, but it's just not feasible on a mobile device. At least not yet. Do you really want your phone's mic on 24/7 feeding audio to the processor, constantly running voice recognition on every sound it hears? Imagine how horrible the battery life would be. Not to mention the privacy concerns of that (and legal issues, I can see something like this breaking wiretap laws in several states).
Even Google Glass requires you to either touch it or activate it through a head motion so it starts listening. The same with Siri on iOS. The point is you only have to do one thing to start issuing a variety of commands. You can do it without looking at the phone.
rhd-android said:
Wow - this is significantly less impressive than I had envisioned. I was under the mistaken impression that Google Now (and Siri on the iPhone) just responded to their trigger words whenever uttered.
This strikes me as somewhat limited in usefulness until they get these apps to listen constantly. If I need to know the weather, why use my hands to launch Google Now, then ask for the weather, and then wait for it, if I could instead just use my hands to launch my weather app directly?
Click to expand...
Click to collapse
As mentioned, think logically about what you are asking for. If your phone's microphone was always on, listening to and parsing every single noise it hears 24/7 to detect the word Google, think about what the battery implications. Every sound it hears it must analyze, all day long. Then, consider any time it thinks it hears Google, it will immediately turn on and listen for more input and then try to respond to that. It's just not feasible to expect this without killing your battery and also dealing with the many false positives that might constantly be making your phone do stuff you didn't intend.
Having said that, you are missing the key part of Google Now which sort of addresses your issues with it. Google Now doesn't listen 24/7 for you to speak to it, it analyzes data such as your search history, location, Google now settings, gmail, calendar, and so on, to proactively give you information before you even ask for it. That's the appeal and the main idea of Now. You still get the searching, the voice commands, etc, but you also get an "assistant" that doesn't need you to constantly tell it what you want. (In theory of course)
jsmith8858 said:
As mentioned, think logically about what you are asking for. If your phone's microphone was always on, listening to and parsing every single noise it hears 24/7 to detect the word Google, think about what the battery implications. Every sound it hears it must analyze, all day long. Then, consider any time it thinks it hears Google, it will immediately turn on and listen for more input and then try to respond to that. It's just not feasible to expect this without killing your battery and also dealing with the many false positives that might constantly be making your phone do stuff you didn't intend.
Click to expand...
Click to collapse
Yep - and that's where I thought the tech was at. I thought we had overcome these problems. I thought we were able to accomplish the above.
I played with it a bunch yesterday, and there was nothing that I could do faster with voice (given that I had to use hands to trigger the app first) than I could do with just hands alone. I tested a lot of common tasks, and launching the app + giving a voice command and waiting for the response action was always slower (not even factoring in voice errors - that's assuming 100% accuracy).
The ability to constantly listen is fundamental to voice commands being useful on a smartphone (in my mind). I just jumped the gun and thought we were there already (adverts certainly make it seem like we are). Until we hit that point, until we have a system that is responsive like KIT, or HAL, or LCARS, I think voice commands are impractical. It needs constant listening. Interestingly, it looks like the industry agrees:
http://www.technologyreview.com/new...d-respond-to-your-voice-even-when-its-asleep/
http://www.techradar.com/news/compu...mm-processor-will-always-be-listening-1132647
^ that tech largely addresses jsmith8858's concerns.
rhd-android said:
Yep - and that's where I thought the tech was at. I thought we had overcome these problems. I thought we were able to accomplish the above.
I played with it a bunch yesterday, and there was nothing that I could do faster with voice (given that I had to use hands to trigger the app first) than I could do with just hands alone. I tested a lot of common tasks, and launching the app + giving a voice command and waiting for the response action was always slower (not even factoring in voice errors - that's assuming 100% accuracy).
The ability to constantly listen is fundamental to voice commands being useful on a smartphone (in my mind). I just jumped the gun and thought we were there already (adverts certainly make it seem like we are). Until we hit that point, until we have a system that is responsive like KIT, or HAL, or LCARS, I think voice commands are impractical. It needs constant listening. Interestingly, it looks like the industry agrees:
http://www.technologyreview.com/new...d-respond-to-your-voice-even-when-its-asleep/
http://www.techradar.com/news/compu...mm-processor-will-always-be-listening-1132647
^ that tech largely addresses jsmith8858's concerns.
Click to expand...
Click to collapse
It's not there yet. It is still very handy when you're multitasking and the best example is driving. Sure I would love to not have to tap my phone a couple if times, but it is still worlds better than losing focus taking my phone out if dock, making those couple taps and typing out my message or commands. Not to mention keeping a CPU and microphone running at all times keeping your phone awake at the same time and it is just not viable. I don't think apple could pull it off either as it would just keep iOS running and running when the phone is not in use. Then there's the privacy factor. I don't think this will be viable for at least another couple of years.
Sent from my GT-N7100
Siri also doesn't respond by keyword launch. S-voice seems like the only one that does and it's definitely less than stellar. As jsmith8858 said above, it's a proactive assistant. Google Now is there for you to give you information it believes you might need before you ask for it. I.e. If i google a place on google maps on my desktop, When i look at google now on my phone a moment later, it has the pop up of directions and estimated time (that i can bring up in maps or navigation by tapping) of where it was when I just searched.
There are third party apps you can use with custom roms that you can bind double-tap home button to initiate voice search, or you can add voice search as a shortcut on your lock screen as well.
I was watching TV and someone on TV said Google, and it triggered the app so I'm pretty sure that there is a way to do it without using your hands, especially because of how many times they said there was a way during one of those recent Key notes regarding it
KILLplay said:
I was watching TV and someone on TV said Google, and it triggered the app so I'm pretty sure that there is a way to do it without using your hands, especially because of how many times they said there was a way during one of those recent Key notes regarding it
Click to expand...
Click to collapse
As far as I know, the only phone that triggers Google now by saying "Ok Google now" is the new moto X, but it has a processor dedicated to ear what you are saying.
I don't know if this ability cam be ported to our s4 and what will be the price to pay about battery consumption.
Enviado desde mi SGH-I337 usando Tapatalk 2
I'm still trying to work up a way to get a decent TTS voice engine on the NST, but what I have come up with so far may be of some use for Glow4 users who wish to try out TTS in reader applications with the Bluetooth capability.
First, SVOX, which was my primary target (because the voices are many and really, really good) looks hopeless. The engine installs OK and works all the way down to Android 1.6, but the voices--even the free "trials"--require PlayStore access to complete installation and SVOX is long gone from the PlayStore, so the whole thing just hangs. I've tried various dubious schemes provided by some users, including the full set of voice data, but they don't work because the app can't be made to see that the data is already present. At least that's my experience.
But Ivona! Also gone from the PlayStore, also works down to at least Android 2.1, and I just tried it on KitKat and it works very well there.
Here is a link to what is called a "complete" collection of voices as well as the engine: https://www.reddit.com/r/tasker/comments/dnspl3/tts_ivona_tts_engine_collection_all_voices/. Follow the directions for installation.
Alas, there are too few voices, and both English voices are female. I have a preference for male readers (please don't flame me...I have heard some very well-read audiobooks done by women) which is why I initially struggled with SVOX. That said, the Kendra voice is light-years ahead of the default Pico TTS voice or even some of the early Google TTS voices.
On KitKat, the change in engines is easy. Go to your Accessibiliy settings (well, if this is non-existent on a Glow4, someone will have to be creative...) and select Ivona, then configure the voice. This should result in a system-wide change. To test, I installed AlReader and opened a book. When I opted for TTS, Kendra started right up.
Alas, for the NST/G the story does not yet have a happy ending. The tts-extended service app does allow you to "change" the default engine, setup the default voice, configure it, etc. BUT, the vast majority of apps ignore these settings and use the Pico TTS engine and the dreadful voices included with that. According to the SVOX FAQ site:
Several apps support changing voices on Android 2.1: Talk To Me Classic, PDF to Speech, Smarter Alarm, Google Translate, and others. These apps use the Extended TTS API. Apps that don’t support the Extended TTS API, such as Google Navigation, only let you change the voice on Android 2.2 and later.
Click to expand...
Click to collapse
I take it from that there must be magic code which is not present in, for example, AlReader, which would allow it to use the Ivona voice. Apparently this problem went away with Android 2.2. Naturally.
So...I'm still poking around for a work-around.
Edit: Eureka! See post #2 below!
[NST/G] AlReader speaks!
I have finally managed to modify AlReader 1.77 to work with the Extended TTS API. This allows the settings app to select the default voice engine and voice, rather than simply defaulting to Pico TTS (which is what otherwise happens in Android 2.1). Having said that, I can't claim it's always pretty. It seems like the much more complex voice data puts a bit of a strain on the NST/G. Once it gets up a head of steam, it's not bad, but it is a little slow to start. It "may" help to play a section of a book using Pico or even espeak (see below) and then exit the reader, go back to the settings app and reselect the voice engine for one of the Ivona voices. I'm not quite sure yet, still playing around. Curious? There's a bit of work to do. Here's how:
1. Install svox and espeak voice data to your sdcard. Here is a direct link to the Eyes Free project code repository: https://storage.googleapis.com/goog...gle.com/eyes-free/voicedata-(espeak+svox).zip.
Unzip the file and copy the two folders of voice data to your sdcard. Without these files present, the next step will hang because it tries to download this data from the Android Market.
2. Install the TTS Service Extended app from the Eyes Free Project repository. Here's the link: https://storage.googleapis.com/goog.../code.google.com/eyes-free/tts_3.1_market.apk. This is essentially an Accessibility Settings app for TTS which includes a pre-Google Voice api to enable voice engine selection for apps that use the api. It also installs the espeak voice engine (worse than Pico, really).
If you run the TTS Service Extended app you will find the espeak voice engine as an option (on the second screen of settings). Pico is the default.
3. Download and install the Ivona voice engine and any voices you want as described in the first post. I have tried both Kendra and Amy. Localization doesn't seem to be an issue for this business, so you can select UK English even if your device is localized for the US. Or, there are a few other English voice options (Welsh, Australian). Knock yourself out!
4. Open the TTS Service Extended app and go to the end where you can select the Ivona voice engine, then configure the voice. Back on the first screen you can set the default engine (Ivona) and select the Language to use. Don't check the box to have this app override the settings of whatever app is using the TTS service.
5. Install the modified AlReader 1.77 (attached below). If you currently use AlReader, you will lose all your settings unless there is a way to back them up and even then they might not transfer since this modified app has a new signature. Not sure about that. You cannot install this modified version over your current version, but must uninstall the old version first. Caveat emptor, blah-blah-blah.
I never did find a way to get the TTS option into the slider menu system on AlReader, so I ended up assigning it to a tap on the header. But you can also use one of the 9 action tap zones to assign a spot to TTS.
I'd reboot. If you have some trouble initially, try changing the default engine back to Pico or even espeak, let it read a bit and then go back to Ivona. That may sort it. In the end, there may be a buffering issue because of the size of the Ivona language files. It's a miracle that I actually figured out how to do this by examining the smali files of the decompiled app and comparing them with the smali files of the TTS Service Extended app. Some pattern matching and lucky guesses eventually bore fruit.
I'm working on the FBReader TTS plugin next. It seems like it should be simpler, but it's not
Hi, I installed Ivona succesfully but today Amy stopped working...I saw a message on the app saying "This is not downloaded from Play Store" French is still working..i tried everything (install, reinstall, restart...) so I tried to follow your instructions to install the tts extended, but I don't find where to extract the first two folders...I would appreciate if you make it clear, Amy was amazing...
Thanks!
josh90sp said:
Hi, I installed Ivona succesfully but today Amy stopped working...I saw a message on the app saying "This is not downloaded from Play Store" French is still working..i tried everything (install, reinstall, restart...) so I tried to follow your instructions to install the tts extended, but I don't find where to extract the first two folders...I would appreciate if you make it clear, Amy was amazing...
Thanks!
Click to expand...
Click to collapse
Just so I am clear before I start in on this, are you talking about the NST? It sounds like you did not at first have the TTS Extended app installed. I don't see how the Ivona voices could work at all without that on the NST.
I was exprimenting with responsivevoice.org - with tablet other than nook
Idea was to send text to his server as a HTML with settings and get it back as an audio stream.
Now they need to create an account to get access - but using it will be free.
Ivona
"Just so I am clear before I start in on this, are you talking about the NST? It sounds like you did not at first have the TTS Extended app installed. I don't see how the Ivona voices could work at all without that on the NST"
Yes, I downloaded the ivona apk(one for each language) and then installed also the Ivona Text-to-Speech HQ apk. After, i placed the audio folder into Android/obb/ and opened the different applications, and signed the learning agreement.
So i could choose the Ivona engine from Settings menu and it worked, I was using Moon Reader and I used it for several days until yesterday, when the unpleasant Samsung english voice surprised me.
Since then, I am only allowd to listen in french voice, that's how I arrived here. I got the apk you were talking about and the two folders but I dont know where to put them in order to get the app to work. It says: downloading from play store and then it closes.
Thanks for your reply!
josh90sp said:
Yes, I downloaded the ivona apk(one for each language) and then installed also the Ivona Text-to-Speech HQ apk. After, i placed the audio folder into Android/obb/ and opened the different applications, and signed the learning agreement.
So i could choose the Ivona engine from Settings menu and it worked, I was using Moon Reader and I used it for several days until yesterday, when the unpleasant Samsung english voice surprised me.
Since then, I am only allowd to listen in french voice, that's how I arrived here. I got the apk you were talking about and the two folders but I dont know where to put them in order to get the app to work. It says: downloading from play store and then it closes.
Click to expand...
Click to collapse
I cannot reproduce what you are describing. I don't understand your comment about the "Samsung" voice. On the Nook Simple Touch there are only the Pico TTS voices possible (and only if you install the voice data--svox at a minimum). The only way to get anything else is to install the TTS Extended settings app and find (or create) an application that supports this interim step toward Google TTS (which arrived with Android 2.2). That's what I describe in the second post. First the svox voices must be installed. Then the TTS Extended app. Then the Ivona engine app is installed. Then the Ivona voice data is copied into the Android obb folder as described.
After this, the Ivona voice engine is selected in the TTS Extended app and a particular voice selected. Now, I have only used this a little and not over a period of several days, so maybe there is a trigger somewhere based on time, but my understanding of these voices is that even when available on the Market, they were free, so I'm not sure why there would be a time limit.
My NSTG where I have this installed does have rudimentary GApps. I'm not sure if that matters.
nmyshkin said:
I cannot reproduce what you are describing. I don't understand your comment about the "Samsung" voice. On the Nook Simple Touch there are only the Pico TTS voices possible (and only if you install the voice data--svox at a minimum). The only way to get anything else is to install the TTS Extended settings app and find (or create) an application that supports this interim step toward Google TTS (which arrived with Android 2.2). That's what I describe in the second post. First the svox voices must be installed. Then the TTS Extended app. Then the Ivona engine app is installed. Then the Ivona voice data is copied into the Android obb folder as described.
After this, the Ivona voice engine is selected in the TTS Extended app and a particular voice selected. Now, I have only used this a little and not over a period of several days, so maybe there is a trigger somewhere based on time, but my understanding of these voices is that even when available on the Market, they were free, so I'm not sure why there would be a time limit.
My NSTG where I have this installed does have rudimentary GApps. I'm not sure if that matters.
Click to expand...
Click to collapse
Thanks for your reply. Sorry if I wasn't clear, and I think I jumped into the wrong post. I am currently using a smartphone, I was looking for any help for my problem.
I will try to reset the phone because I was very happy with those voices.
Thanks again!
I'm building a WearOS app and would want to utilize on device speech to text convertor. Is there an example or api documentation I can refer. My search ends up with Gboard, which is not what I am looking for.
The app is something like virtual assistant and listens to user's voice.
You can try mp3 to text converter. It is cheap, fast, and easy to use. Also, you can choose between Automatic and Manual Transcription Services. It depends on your needs, budget, and time.
I registered finally! been reading this site for years. (for like ten phones worth).
I just wanted to ask, if the app is/went well? You did not get good response to your question, and well, that bummed me out. (nice try McLellan, but too little too late, truly).
I would assume you would want to utilize Google's speech to text engine, to get the text, and then script out what to do with the input yourself. I AM look for an opensource solution to G-S-to-T right now, myself....
-JJ
Yeah OK, 2019 was a triumphant year for people that use their voice to a keyboard: https://venturebeat.com/2019/08/16/google-open-sources-live-transcribes-speech-engine/
now I just gotta figure out why AICP is not, using it.....
So, I fixed my problem. Gboard, for whatever reason, had rendered itself useless. So using Aurora, I uninstalled, re-installed , configured, and now i can talk out my SMS text messages in public to all-who-can-hear-me's annoyance. And, I can still talk to the voices in my head and look sane too (and my phone don't even need to be on! just in my hand.) I have never tried WearOS, but, I might need to look that up, cuz I keep hearing about it. Good luck to you and such.
-JJ