New Feature Request - Link to Speech to Text capability

[FairfieldPhoto]

I am beginning to run out of new feature ideas for PM, but here is one that may give you something to chew on for awhile.

I use the voice recorder on my camera a LOT.  From capturing competitor numbers before they compete to noting which event is underway to special notes that will go into a caption, my voice recorder gets a lot of use.

I also know that there is a lot of innovation going on between Google, Dragon, and others to create reliable speech-to-text capability

Would there be a way to create a variable where that when encountered, would take any associated WAV file and send it off for conversion to text? Something like {speech2text} / {s2t} and what it would return would be the text string the conversion algorithm came up with.  I am not sure that the speech conversion is something that would need to exist wholely within PM -- it would drastically inflate the size of the app plus with so much innovation going on, whatever would be created would be obsolete soon after.  It would be better if PM used an Internet connection to do that conversion against some kind of service in the "cloud".

Would anyone else find this useful?

-Mike

[Sven]

Hi Mike!

Sounds interesting. I was searching for a software with "OCR"-capability.
At sport-events the competitors wear their bibs with numbers on it.
Would be great to have "a software" to detect the numbers and then going on with the stuff you mentioned... (yes, i know that this is really CPU intense...)
Often there is no time to record some notes...

But the idea is nice in my opinion.

For people who do not need this kind of feature: Would it be an option to create a "plug in"-feature in PM?
Have the base-part of PM and the option to purchase features, not needed by others?

BR
Sven

[vAfotoriporter]

Both ideas are very interresting and could be great additions to PM. I don't know much about OCR or speach to text technologies but I think for this Camera Bits should have to develop their own OCR and Speach to Text methods what would greatly tie their resources.

I think the closest and perhaps fastest solution would be to find some software that could produce txt outputs from the images and voice recordings. The OCR/speach2text program should generate a tab separated text file containing image filename and visual/voice info contained in the image/recording.

Then this text file could be easily used as a code replacement file in PM with its already existing code replacement features. Using already existing features of both sides could eleminate the need for modifications. This might work without the need for any modification either on the PM end or the OCR/speach2text software end.