Active Topics
-
[ANNOUNCE] MeeShop - new app store (10)
to Applications by aspergerguy - 1 day, 8 hrs ago -
HTTP-401 Sfos-Builder when using /etc/ssu credentials (0)
to SailfishOS by insanelysexy - 2 days, 14 hrs ago -
Longtime users: What device(s) are you using today? (54)
to General by ryu1 - 4 days, 19 hrs ago - more...
skate_boy92 |
2013-02-26
, 10:48
|
|
Posts: 121 |
Thanked: 60 times |
Joined on Jun 2012
|
#611
|
Install via FAP in N900, it says Failed to fetch packages - Your network connection might be down or your catalogs could be out of date
| The Following User Says Thank You to skate_boy92 For This Useful Post: | ||
|
|
2013-02-26
, 11:27
|
|
Posts: 121 |
Thanked: 60 times |
Joined on Jun 2012
|
#612
|
Finally, it was installed
But It doesn't start
Edit: After reboot, it works
How to use front camera?
Last edited by skate_boy92; 2013-02-26 at 11:30.
But It doesn't start
Edit: After reboot, it works
How to use front camera?
Last edited by skate_boy92; 2013-02-26 at 11:30.
| The Following User Says Thank You to skate_boy92 For This Useful Post: | ||
|
|
2013-03-16
, 00:32
|
|
Posts: 669 |
Thanked: 433 times |
Joined on May 2010
|
#614
|
taixzo, is it possible to make a script that listens to mic and prints to output what it heard?
i just thought that this way we'll be able to use scripts to do generic voice recognition to whatever usage anyone likes.
for example, i can record my voice saying "car", get the output recognized text and use it to activate my car mode (even if it was not recognized as "car")
such customization will be awesome and there will be no need for a GUI application for many users.
what do you think?
i just thought that this way we'll be able to use scripts to do generic voice recognition to whatever usage anyone likes.
for example, i can record my voice saying "car", get the output recognized text and use it to activate my car mode (even if it was not recognized as "car")
such customization will be awesome and there will be no need for a GUI application for many users.
what do you think?
| The Following 2 Users Say Thank You to impeham For This Useful Post: | ||
|
|
2013-03-18
, 16:34
|
|
Posts: 959 |
Thanked: 3,427 times |
Joined on Apr 2012
|
#616
|
Originally Posted by impeham
You can; however you'd probably be better off starting off with the example Pocketsphinx script, as I did with Saera, rather than trying to adapt the code.
taixzo, is it possible to make a script that listens to mic and prints to output what it heard?
i just thought that this way we'll be able to use scripts to do generic voice recognition to whatever usage anyone likes.
for example, i can record my voice saying "car", get the output recognized text and use it to activate my car mode (even if it was not recognized as "car")
such customization will be awesome and there will be no need for a GUI application for many users.
what do you think?
All:
I haven't posted any updates on Saera for a while, because I've been working on a component that doesn't really allow partial updates: the A.I.
The current system uses hardcoded functions for everything, which isn't very flexible. Consider the following: You have asked for directions home, and then on the way say: "Will I reach my destination before the Giants game?" What function would handle this: the function that deals with directions? The function that returns sports information? It's not feasible to pack every conceivable scenario into each function.
However, there's a different approach: Use a node-based system and allow the program to find the connections between nodes. That statement would be split up into nodes: the "destination" node, the "event information" node, the "before" node, and the "will" node which is the root node. The "destination" node would have an "ETA" property, which can be connected to an "ETA-to-time" node which has a "time" property; that "time" property and the "time" property would connect to the "before" node, which has a Boolean property (true or false) that would connect to the "Will" node and return an answer.
This system is still in development and has not been integrated with Saera yet, but I figured I would post an update.
__________________
Saera: A Siri clone for the N900, N9(50) and Jolla
Follow me on Twitter: twitter.com/taixzo
Saera: A Siri clone for the N900, N9(50) and Jolla
Follow me on Twitter: twitter.com/taixzo
| The Following 28 Users Say Thank You to taixzo For This Useful Post: | ||
|
|
2013-03-18
, 17:13
|
|
Posts: 951 |
Thanked: 2,344 times |
Joined on Jan 2012
@ UK
|
#617
|
Originally Posted by taixzo
Really nice to hear your working on some really cool stuff
You can; however you'd probably be better off starting off with the example Pocketsphinx script, as I did with Saera, rather than trying to adapt the code.
All:
I haven't posted any updates on Saera for a while, because I've been working on a component that doesn't really allow partial updates: the A.I.
The current system uses hardcoded functions for everything, which isn't very flexible. Consider the following: You have asked for directions home, and then on the way say: "Will I reach my destination before the Giants game?" What function would handle this: the function that deals with directions? The function that returns sports information? It's not feasible to pack every conceivable scenario into each function.
However, there's a different approach: Use a node-based system and allow the program to find the connections between nodes. That statement would be split up into nodes: the "destination" node, the "event information" node, the "before" node, and the "will" node which is the root node. The "destination" node would have an "ETA" property, which can be connected to an "ETA-to-time" node which has a "time" property; that "time" property and the "time" property would connect to the "before" node, which has a Boolean property (true or false) that would connect to the "Will" node and return an answer.
This system is still in development and has not been integrated with Saera yet, but I figured I would post an update.
I wish you best of luck
 |
|
2013-03-19
, 10:19
|
|
Posts: 5,028 |
Thanked: 8,613 times |
Joined on Mar 2011
|
#618
|
taixo, AI development is exactly what I hoped for, no matter how long it will get to appear. It is what makes this project ingenious, and make it much more interesting than "original" siri -the latter is just fancy voice UI, where Saera becomes real AI.
/Estel
/Estel
__________________
N900's aluminum backcover / body replacement
-
N900's HDMI-Out
-
Camera cover MOD
-
Measure battery's real capacity on-device
-
TrueCrypt 7.1 | ereswap | bnf
-
Hardware's mods research is costly. To support my work, please consider donating. Thank You!
N900's aluminum backcover / body replacement
-
N900's HDMI-Out
-
Camera cover MOD
-
Measure battery's real capacity on-device
-
TrueCrypt 7.1 | ereswap | bnf
-
Hardware's mods research is costly. To support my work, please consider donating. Thank You!
|
|
2013-03-20
, 16:21
|
|
Posts: 359 |
Thanked: 322 times |
Joined on Jun 2010
|
#619
|
Interesting app. It seems like we need (1) a simple animation to show that Saera is working on a request and (2) a button to cancel the last request and try a new request.
Thanks for your efforts! :-)
Thanks for your efforts! :-)
__________________
Hitting this thanks button will make you smile and feel good ----------------------------------------------------------------->
Hitting this thanks button will make you smile and feel good ----------------------------------------------------------------->
|
|
2013-04-16
, 15:11
|
|
Posts: 29 |
Thanked: 77 times |
Joined on Mar 2013
@ Australia Maaaaaaaaate
|
#620
|
Taixo.What are the chances of an update on your development ?.Its been all too quiet,which it seems is unusual for you....
| The Following User Says Thank You to ghost_who_walks For This Useful Post: | ||
 |
| Tags |
| saera, speech-to-text |
«
Previous Thread
|
Next Thread
»
|
All times are GMT. The time now is 03:43.